'iiii DEl'AKT-MKN'l' (»F THE INTERIOR M0N()(iHAPH8 OF THE United States Geological Survey YOLUAiE XLVIII I>Mit I- TEXT WASHINGTON GOVERNMENT PRINTING OFFICE 1 !t 5 UNITED STATES GEOLo* !I( \\L SUTJVEY CHARLES D. WALCdTT, DinEcruK s r A T u s l)K THE MESOZOIC FLORAS OF TUF UXITFI) STATES I'.Y LESTER F. )VAKD WITH THK ri) 1j I, ABO U A TIOX l)F William A1. Fontaink, AiiTHtiu Bikbins, and (i. \{. Wieland Pf^irt I.-TKXT .-IBRARY NEW YORK BOTANICAL OARDEN. WASHINGTON GOVEKXMENT PRINTING OFFICE 19 05 p. ( .1- CONTENTS Pagp. Letter of Transmittal, by C. W. Hayes Part I. The Triassic flora (continued) ''^ The ( )l(lcr Mesozoic of Arizona Si lilt isiaphical relations of the Oldei' Mesozoic deposits of Arizona "> 18 Tlie Moeneopie formation i() The Shinaniinp formation The Lithodendron member , 0') The Lcroux member 97 The Painted Desert fornuition Paleontological rehitions 99 The Moeneopie formation 30 Tlie Shinanimp formation - The Painted Desert formation "'' a ,. 37 Sections Section I. Canyon of the Little Colorado 3' Section IL Moeneopie TV ash Section IIL The lower LittU' Colorado Valley ^^ * 40 Section IV. Winslow • 42 Section V. Leroux Wash ■ • Section VL The Petrified Forest '*'^ Section VIL Red Butte '^^ JO Section VIIL The geological column Part II. The Jokassic flora (continued) 47 The Jurassic flora of Oregon The Jurassic flora of Douglas County, Oreg., by William M. Fontaine 48 Descriptions of the species ■'' Other plant-bearing beds in the Jurassic, or forming the transition to the Lower Cretaceous 14,5 Report of Professor Fontaine on the collections from the beds above named 1-18 1 . Plants from Curry County, Oreg ^'^f 2. Plants from Herendeen Bay, Alaska ^^ 3. Plants from the Copper River region, A!a.ska l''^" C3 4. Plants from the vicinity of Cape Lisburne, Alaska J5.3 22 5. Plants from northern California I 6. Plants from northern Montana lO 7. Plants from Slate Springs, Monterey coast, California I7(i 177 178 6 CONTEXTS. Part II. — The Jirassic flora — Continued. Pnge. Jurassic <'ycaia- Continued. Fossil plants from the new reservoir -^l" Fossil plants from Terra Cotta -^l" Fossil plants from Ivy City '^'^ Fo.ssil plants from Langdoii ■'''^ General remarks '^"'^ Fossil plants froTu the Queens Chapel road •''-' Localities in Maryland Fossil plants from Hosiers Bliilf Fo.ssil plants from Kiverdale Fossil plants from near Bern'yn « ^^^ Fossil plants from the Bewley estate ^^ Fossil plants from Mnirkirk 0^4 Fossil plants from Contee '•'' Fossil plants from Arlington ■"'■" Age of the Arlington heds -^-^^ Fossil plants from Hanover °^'^ Fossil plants from the Howard Brown estate 544 Fossil plants from Reynolds's ore pit •''44 Fossil plants from German's iron mine •'45 Fossil plants from Hobbs's iron mine '^'^ Fossil plants from Tiptop •''"'■5 Fossil plants from Vinegar Hill •''4' Fossil plants from Soper Hall •~'^^ Fossil plants from Lansdowne •'■'''' Fossil plants from Federal Hill ^^^ Age of the Federal Hillbeds •'i<56 Fossil plants from Union tunnel •'^''0 Fossil plants from Springfield •''"^ Fossil plants from Stemmers Run . •"'"! Fossil plants from Broad Creek ■'''2 Fossil plants from Plum Creek -WS Fossil plants from Muddy Creek ■" 2 Fossil plants from Locust or Poplar Point -"S Fossil plants from Grays Hill •'•"•^ General remarks and conclusions -"4 Correlation of the Potomac formation in Virginia and Maryland -^^O Table of distribution of Potomac plants 582-589 Analysis of the table ^^ Columnar .section of the Potomac formation •"'■'' •'■''^ Index ^^ ILLUSTIIATIONS, FHIURKS. IN l'\UT I. Page. Flo. 1. Sccti.iii ..f till' ciiiivon of lb.' T.iltlc C.ilniailu. Aiizoiiu •'^'^ 2. SiM'liciM i>f Ihi' MoiMiciipio 'Wash, .\iizoiui 3. Section of the lower Little C'oloniilo Viilley, Arizona 4. Section tlirou<;li Winslow, Arizona 5. Section near Lenaix Wash. Arizona JO 6. Section throuf;h the Petrified Forest of Arizona • - 7. Section of Rcil Butte S. Geolosii-al cohinni of the Older Nh'sozoic of Arizona ^•' 9. Section of Cow Creek, Nickel Mountain, and Buck Peak, Douglas County, Oreg -17 10. Potomac exposure on Sixteenth street, Washington, D. C '^' ' 598 1 1 . Cohunnar section of tlie Potomac formation PLATES, TN P.\RT IT. p,^_,,.. Coniferous plants from the Trias of Arizona '"" ' Sketch map of the Little Colorado Valley, Arizona, and adjacent regions ''*' Sketch map of the Buck Mountain region and Cow Creek Valley, Douglas County, Oreg v Jurassic liverworts and ferns from Oregon ^'' Jurassic ferns from Oregon Jura.ssic ferns and Equiseta from Oregon ^^ Jurassic cycads from Oregon Jurassic cycads and William.sonias from Oregon ^^'^ Jurassic Gingkos from Oregon -^ ' ' ' ' Jurassic Gingkoacese and Taxacese from Oregon xxxiv Jurassic conifers from Oregon '^^- ^' - • • Miscellaneous Jurassic plants from Oregon xxwii Jurasso-Cretaceous plants from Oregon and Alaska xxxviii Jura-sso-Cretaceous ferns from Cape Lisburne, Alaska xxxix-xi.in Jura.sso-Cretaceous cycads and Gingkoaceae from Cape Lisburne, Alaska x"->v Jurasso-Cretaceous conifers from Alaska, Montana, and California XLV 9 10 ir>r.rsTHATioxs. Platf". Jurassic cycnds from Wvoniing xi.vi-i.xiii Sketch map showing fossil localities of the Shasta IdirnalKin oft iililorniu i.xiv Ferns from the Shasta formation of California and Oregon Lxv-LXVI Cvcads from the Shasta formation of California and Oregon i.xvii Cycads and conifers from the Shasta formation of California and t)rcj;on i.xvni Conifei-s any mc from November 3 to 16, 1899, an account of which was given in the first paper (pp. 320-332), only sufficed to indicate in a general way the great interest that attaches to the region visited and the possibilities it possesses from both the f'The first paper appeared in Twentieth- Ann. Rep. U. S. Geol. Survey, Pt. 11, IWHt, pp. 211-748, pis xxi-clxxix. U MESOZOIC FLORAS OF rNITED STATES. stiatijirapliical and the palconlolopcal points of vicnv. It was my intention at that tinu' to improvo the first opportunity that presented itself to conduct a much more extensive campaign into that region. 1 found tlu> country so exceedingly dry in November that 1 imagined, and indeed was told by persons who live there, that sj^ring would ])e a much more advantageous season for such a campaign. 1 therefore decided to make a somewhat exhaustive study of this region, with an appropriate outfit, in the months of May and .lune, l!)()l. The discovery that I made on November 14, 1899, of fossil bones neai- Tanners Crossing of the Little Colorado (see first paper, pp. 322-323) seemed to make it one of the prime requisites of such an expedition that it !;)e accompanied by a competent vertebrate paleontologist, well versed in the methods of collecting and preserving fossil bones. When the attention of Prof. H. V. Osborn was called to this subject he manifested a lively interest in it and offered to select a suitable person to accompany me for this purpose. The choice fell upon Mr. Barnum Brown, of the American Museum of Natural History, wdiose success as a collector of fossil vertebrates and as a field naturalist in general has secured for him wide recognition. The United States National Museum, being in need of a series of Triassic l)ones, assumed the respon- sil)ility for this part of the expedition. The rendezvous was at Hol- In-ook, Ariz., on May 7, where an adequate outfit was procured and the expedition started on the 8th. The route followed was nearly the same as that pursued by me in 1899. The object was to make as complete a study as possiljle of the geolog}^ and paleontology of the Little Colorado Valley, as it is in that \-alley, or at least in that drainage, that nearly all the older Mesozoic of this part of Arizona is to be found. Certain signs of the occurrence of Triassic vertebrates made them- selves known at different points along the route, but only as weathered out on the surface in such a manner that their original source could not he traced. But as soon as we reached the region of vai-iegated marl buttes, some 5 or G miles above the Lees Ferry road, such bones l)egan to be found and their position in the beds located. Nearly three weeks' careful investigation of the entire region in which such beds occur proved that the small group of buttes in which I first found the TIIH OLDER MESOZOIC OK ARIZONA. 15 t)()iies was the I'ichcsl in this respect, and the greater i)art of Mi-, lii'own's collection was made within a mile ' to be foimd mainly in the great abundance of gypsum that prevails throughout all these beds, which, as is well known, is almost fatal to the preservation of organic remains. But for it no doubt the vertebrate remains would be much more abundant and complete, and it is onl>- the silicified wood that seems to escape its influence. This latter, however, is mainly deposited in beds of sand, coarse gravel, or conglomer- ate, which, in themselves, ar(> unfavorable to plant impressions. "A prclinuimiy rejiort was piiMislicil hy liim ill Science, \. S., Vol. XIV, September 6. H)l)l, p. 370. Ifi JIESOZOIC FLORAS OF FNTTED STATES. The al)senc'e of plant impressions enabled nu> to (Unote more attention to geological considerations than would have hocn })racticabl(" had large collections of plants been made; and I regard the geological results as of sufficient importance to Ite introduced here somewhat full\-. This chapter will therefore be divided into two parts, the first of which will deal with the stratigraphy and the second with the paleontology. STRATIGRAPHICAL RELATIONS OF THE OLDER MESOZOIC DEPOSITS OF ARIZONA. The geology of the (irand Canyon region of northern Arizona has received much attention on the part of geologists, and considerable has been written on the higher Ijeds of Mesozoic age that lie to the eastward and northward, Ijut very little study seems to ha^e been made of the Little Colorado Valley above Coconino Point, where it broadens out into a plain. The strata of the Grand Canyon up to and including the junction of the Little Colorado with the Colorado River consist, as all know, entirely of Paleozoic and pre-Paleozoic rocks, and it is the Carboniferous limestones, or sometimes sandstones (LTpper Aul)i'ey). that occupy the surface of l)oth the Colorado and the Kaibab plateaus. But the entire system dips sensibly to the northeast, and at any point some distance back from the canyon remnants of Mesozoic rocks occur for many miles to the west of the Little Colorado. That river, therefore, practically flows for almost its entire length over Mesozoic strata, but these do not attain their great development except on the northeastern slope of the valley. Here they form se^•el■al series of terraces, rising one above another back- ward from the river, and forming at their maximum de^•elopment lofty and picturesque escarpments, with brilhantly colored stratification, rival- ing in many respects the Grand Canyon itself. The broad, arid plains that lie to the southwest of these cliffs have received the name Painted Desert, from the circumstance that from any jioinl on this desert these "painted" cliffs are always hi full view. From a great distance they may under certain conditions appear beautifvd and innocent, but any attempt to invade this desert or to scale these cliffs, except by means of the few well-known Indian trails, is certain to be met with defeat, and the hardships that have to be endured in striving to traverse this region are of the severest kind. TllK OLDKli MlLtiOZOIC OF ARIZONA. 17 Vovv little seems to he known of llie more detailed nature of these deposits. They ai'e usually spoken of as a siniil(> <2;reat g;roup of beds, and I am not aware of any serious attempt to subdivide them or arrange them into anything like sucee.s.nve formations. It was my chief object during my Slav in that country to subject tliese deposits to a searching analytic:d study and to woi'k out if possible their ti'ue succ(>ssion. I began this study, as already shown, by a reconnaissanc(> of the Little ('olorado Valley. After making camp at Tanners Crossing, which is only 12 miles above the point where the Little Colorado enters the limestone canyon at the foot of Coconino Point, 1 set about mastering the details of the stratigraphy of that general region. Later on, and in the light of information tliusol)tained, I studied th(» various remnants of the Mesozoic that are scattered over the Colorado Plateau, and especially Red Butte, which is the most conspi(;uous and best known of these remnants. Finall>-, as a concluding task, I returned to the upper portion of the valley of the Little Colorado and made a study of the group there similar to that which I had made below. I shall be obhged to omit a great amount of important data, including many sections recorded in my notebook, and shall give only the most general and essential results and reproduce the general sections that most clearly illustrate the phenomena. I think tliat I have succeeded in dividing the group into three entirely distinct formations. One of these, the thickest of them and the one which is best known, has already been named by Major Powell the Shinarump." This, however, occupies the central portion of the beds in their geological sequence. The other two divisions are, so far as I am aware, unnamed. The lower beds I therefore designate the Moencopie beds, from having first found them in their full development at the mouth of the Moencopie Wash. To the other, or highest formation of the group, I have thought it appropriate to give the name Painted Desert beds.'' "Geology of Ihe Uinta Mountains, etc., 1876, pp. 68-69. Soe Twentietli .\nn. Hop. U. S. Cicol. Survey, Pt. II, 1900, p. 318. ''Tlie name "Painted Desert" occurs, apparently for the first time, in the contents to Chapter IX of Part I of Lieutenant Ives's Report upon the Colorado River of the West, pp. 15 and 113, l)ut is not used in the description of the desert on pp. 116-117. It is used by Doctor Newberry in Part III, on pp. 76-83, and to it lie devotes a section. Tliese early uses of the term show that it refers to an area lying opposite to the region between Wolfs Crossing and Wiuslow, but Doctor Newbeiiy says (j). 76) "that the peculiar ijliysical aspect and .MO.N .xi.viii — 0.5 2 18 MESOZOK" FLORAS OF UNITED STATES. THE MOF.XCnPIE FOlUfATIOX. Those occupy \\\c lowest portion of tiie group, liaving a maximum observed thickness of l^etween 600 and 700 feet. They present several distinct phases, but the greatest part of them consists of dark-reddish lirown, soft, laminated, argillaceous shales, nearl}- destitute of silica, highly charged with salt" and g3'psum, tending on exposure to assume the character of nearly homogeneous marls and to form low ridges, but- tresses, antl even isolated knolls or buttes, at the bases of cliffs and in eroded valleys. The gypsum often forms thin sheets which appear as fine white lines and which do not follow the planes of stratification, but cross the beds irregularly and also cross one another, giving the exposures a peculiar striped appearance. Between these beds of shale there occur, usually at more than one horizon, brown sandstones. These are more or less argillaceous and their geologiciil structure of the Painted Desert prevail over a wide belt of countiv liorderiiig tlie Little Colorado on the east, and e.xtending at least as far northward as our camp 7:-i." Tiiis camp appears from the very imperfect map accompanying the report to have been about on tlie latitude of Tanners Crossing, Ijut far to the westward. On this map the Painted Desert is represented as occupying all that region lying along the southwestern base of the painted cli(Vs from the line of their route through the gap at Blue Peaks and Pottery Hill northwestward to an indefinite distance. On the latest Land OfTice maps, however, it seems to be restricted to that portion of the desert lying north of the Moencopie Wash and along the base of Echo Cliffs. There seems to be no good reason for thus restricting it. "An artesian well was bored at Adamana, on the Santa Fe Pacific Railroad, S miles north of the Petrified Forest and in the valley of the Rio Puerco. At a depth of 305 feet water was struck which had sufficient force to rise 19 feet above the surface and discharge 25 gallons per minute. The water was very salt, reported at 3 per cent chloride of sodium, so as to be whollj- unfit for any use. Mr. James Swainson, in charge of the work, which was done by the American Well Works of iVurora, 111., was good enough to send me the log, which is as follows: Record oj welt boring at Adamana, Ari:. Feet. Surface sand and adol)e 55 Sandstone - 3 Cement gravel 1 Sandstone 29 (Water at 88 feet only slightly salt.) Sandstone 20 Brown shale ■H Red shale 49 Hard brown and blue shale 5 Red shale "0 Sandstone 10 Hard brown shale 20 (Intensely salt water.) Total 305 The lower 200 feet of this section cleaily belong to the Moencopie beds. THE OLDER MESOZOIC OF ARIZONA. 19 exposed faces do not present sharp angles, lail have ronnded forms, due in the main to the influence of winds, wldch wear off the jagged apix'ai- ance but (h) not tend to foi-m chinnieys or assume fantastic shapes. These sandstone ledges, which are very imiform in composition, some- times have a thickness of 100 feet or more, though such h(>avy beds are usually- interrupted by several layers of the shale. Toward the lower part of the Moencopie beds the shales gradually become calcareous, and there is in nearly all good exposures a horizon of wdiite, impure limestone, well laminated in its central portion, but l)ec()m- ing ver>- thin and hard tjelow and finally passing either into the typical shale or into homogeneous marls. The extreme upper and also the extreme lower portions of the Moencopie Ijeds always consist, so far as observed, of the typical dai-k-brown argillaceous shale, and the whole series, wherever the contact can be found, always rests in marked uncon- formity upon the underlying Paleozoic rock (Upper Aubrey). It is very probal^le that the lower portion of the Moencopie beds belongs to the Permian. ' TEE sniNARUMF FORMATION. This constitutes a vast assemblage of strata with a maximum observed thickness of at least 1,600 feet. It presents a number of phases, some of which are so distinct that if studied in only one locality they would naturally be regarded as separate subdivisions, but such a general survey as I have l)een making points to a certain homogeneity in all these beds, or at least establishes the unmistakable tendency toward the recurrence in any of the phases of features that are prominent in other phases. The Shin- arump constitutes the horizon of silicified trunks, and there is no part of it in which fossil wood does not occur in great abundance. It also marks the limit of the wood-bearing deposits of this region. For this reason alone, in view of the etymology of the name, I should be justified in extending the Shinarump as -far as the fossil trunks occur, and it is obvious from the language used that Major Powell had the upper portions of the formation in view as well as the lower when giving the name, althougli other geologists, in speaking of the Shinai'ump, usually seem to have in mind only those beds which he called the Shinarump conglomerate. It is doubtful, however, whether the remainder of the formation has reall\- been studied or carefully observed by others. 20 jMesozoic flohas of umtkd states. I shall divide the Shinanimp into two somewhat distinct parts and call the lower the Lithodench'on member and the upper the Leroux member. THE IJTIIODKXDROX MEMBER. This division is the ec}uivalent of the Shinanimp conglomerate of Powell, anfl I was at first disposed to retain his name under the rule of priority, and did so in my preliminaiy paper," notwithstanding the far greater development and marked change of character which it assumes in t he upper part of the Colorado Valley. Attention has been called to the fact that the use made l\y Major Powell of the name Shinarump conglomerate for a part of a larger group which he called the Shinarump is likely to lead to confusion and is generally objectionable. I shall therefore drop that temi, except as descriptive of the conglomerate beds that occur in the Shinarump, and shall call the part of the formation in which these con- glomerates occur the Lithodendron member. This term is specially appropriate not only because of the stone trees that constitute the most prominent featui'e of the beds, but also because the name was given by Lieutenant Whipple in 1853 to the stream or wash in which petrified tnmks were found in great abundance by his exploring party when it passed through that region.'' This was there called Lithodendron Creek, and it was from the bed of this creek that the two trunks brought to the United States National Museum in 1879 by Lieutenant Hegewald were obtained, these being the only ones that have thus far been studied from the standpoint of their internal structure. The creek lies in the region where the Lithodendron beds attain their maximum development and only a short distance from the Petrified Forest which it has been proposed to set apart as a national park. Although perhaps the most prominent feature of this formation is the so-called conglomerate, which sometimes is in truth deserving of that name, and contains somewhat large but always well-worn pebbles and cobbles derived from imderlying formations, it rarely happens that this aspect of the beds constitutes the major portion of them. In the first place, the conglomerate tends to shade off into coarse gravels and n Geology of the Little Colorado Valley: Am. Joum. Sci., 4th ser., Vol. XII, No. 72, December, 1901, pp 401-41.3. '' Report of Explorations and Siiriey.s to asccrhiin the most Praotiiidjlc and KiDnoniical Konio for a Kailroad from the Mississippi Kivcr lo ihc Pacific Ocean, Vol. Ill, 185(3, Pt. 1, pp. 73-7'r, Pt. U, p. 2S; Pt. IV, pp. 43, 150, 151, 1G7. Till-: OLDIOli .MKS( )/()!(• OF ARIZONA. 21 then into (rue sandstone.s. These sandstones are of a hjilit color, eon- traslinfi strongly with the dark-brown sandstones of (he Moeiicopie ali'eady described. They ai'e, moi'eover, always more or less cro.ss- bedded and usually exhil)il lines of i)ebbles ruiuiing through (hem in various directions. These are (rue sandstones, very hard, de\'oid of alumina, and scarcely affected by the winds, so that their angles are usually sharp and the ledges tliey form are abrupt and jagged. Although the sandstones proper generally occur lower down, there is no uniformity in this arrangement, and sandstones are often found in the middle and conglomerates mon^ rarely at the top. But in addition to these tlie Lithodendron member eml)races other classes of beds. There is a well- stratified layer of thinnish sandstone shales that is often seen imme- diately under the heavy sandstone cap. Some of these shales have a grayish color and are highly argillaceous. These layers tend to thicken even within the meml^er itself, l)ut especially farther out, and, what is more significant, they often become transformed into a bluish white marl. This condition can l)e seen between the beds of conglomerate in places where the Lithodendron beds are comparatively thin, as in the lower valley of the Little Colorado, where they are only about 300 feet in thickness. This feature is not very prominent, l)ut at other places, as in the Petrified Forest region, where the Lithodendron beds attain their maximum thickness of 700 or 800 feet, this tendency on the part of certain lieds to become transformed into marls is the most marked feature of the meml^er. The marls here occupy much more than half of the beds. They are very varied in color, showing jjesides the white and blue tints a great variety of darker ones, such as pink, purple, and buff. These heavy marl beds, of which there may be se\^- eral in the same cliff, are interstratified between conglomerates, coarse gravels, and cross-beddecl sandstones, all of which taken together form the beautifidly banded cliffs that are seen throughout the Petrifietl Forest, especially along its northern flank. It thus becomes necessary to include under one designation all of these varying beds, which often change the one into the other even at the same horizon within short distances. It remains to mention certain minor features, which are not uni- versal, but which nevertlieless have considerable importance. In the •22 MKSOZOIC FLORAS OF L'XrrED STATES. lower Littlo Colorado \'allev (here occur numerous somewhat calcareous (•lav lenses, th(> Vuur takino; the form of brijiht white stripes, while the chiv is usually purple or pink. These are wry distinct objects and vary in size from lenses 10 oi' even 20 feet in lenjith to small lenticular blocks or someuliat o\;il or even spherical cla\' balls oi' pellets. These calcareous clay inclusions are scarcely seen faiihei- to the southeast, ]>ut on Red Butte they are well marked, and hei'c the clay becomes l)rillianl red and constilules a true paint stone. Another fact to l)e noted in connection with ttie Lithodendron beds is that at certain local- ities, notably on Hed l^utte, there is at its base a clear indication of a transition to the Moencojiie formation. The cono:lomerates proper are underlain ))>' arjiillaceous shales closely resembling those of the I\Ioen- copie, l)ut beneath these is a sandstone ledge which can not be referred to the lower ch vision, as it is more or less cross-bedded, possesses con- siderable grit, and has included in it small clay pellets similar to those of tlie true conglomei'ate series, in which I have for this reason included it. This condition of things may be somewhat puzzling from the strati- graphical point of view, but the disadvantage in this respect is much more than compensated for by the evidence that it furnishes in fa\'oi- of the view that all of these beds really constitute one great series, and as opposed to the view which it may l)e inferred that certain geolo- gists hold that the series of these beds which I have included under the name of Moencopie belongs to a different system and is in some way connected with the underlying Paleozoic rocks. This view, in the light of the al)Ove-mentioned facts, is, in my opinion, quite untenable. THE LEKOUX MEMBER. Under the name Leroux I include the remainder of the Shinarump, deriving the name from Leroux Wash," which enters the Colorado Valley 2 miles l)elow Holbrook, and on which, some 15 miles north of Ilolbrook, this member attains the greatest development that I have obsei-ved, probably I'eaching its maximum of 800 feet. These beds, too, if stvuUed at localities where they are less developed, might be sup- "The name "Leroux's fork" was given to this wasli by Lieutenant Wliipplc's pnriy. wlid fdllowed it down some di-stancc and enoiimpcd at its jiincfion with tlie Little Coloiiulo on Dccciiilx'i .'), 1S.53, this Ix'ing their Camp 79. See Pacific Railroad Reports, Vol. UL Ft. I, p. 7.5. The name is written in two words on the Ijiitid Office map of Arizona. TIIK 01,l)i;i; MKSOZOK" OF AIM/OXA. 2o posed to form scvcM'al quite distiiicl sulidh isions. Indeed, 1 \v;is of this opinion dui'in.ii:; most of my s1;iy in tlic lower I/iltle Coloi'ado Valley, but (>\'en hefoi'e lea\'ino; Ihei'e the proofs of their hoinou'eneit y had heeome appai'eni . At l(>ast tli(> lower half eonsists of those remai'kahhOx'ds in which 1 had rrriiiinally found the vc'lehrate bones in IS!)',), and in which .alone thus fai' \'ei'tebrate remains ha\'e been obser\-e(l. 1 lia\(' sometimes desifr- nated ihem the variegated m.ai'ls and sometimes the Belodon beds. The distinguisliing feature of these beds is the jnvsence of '^ivdi num- bers of small buttes, the smaller one.s appearing to be blue clay knolls, l)U( the lai-g(M' ones showing other colors, especially purple, and some- times several bantls of different hues. Almost (>\erywhere at this horizon there exist plains, dotted .all over with tliese remarkable httle buttes, varying from '.] or 4 feet to 20 or 3U feet in height, usually i.solated from one another and luiA'ing a form peculiar to them. They are not conical in the true sense of the word, since they do not rise to a point at the summit, but are always rounded off and have the form of a well-made haystack, the smaller ones looking like haycocks in a field. These butte-studded plains are of course simply the remains of a plateau or mesa which has been worn away, primarily by the action of water, but for a very long period there can be no doubt that wind has been the more potent agency. There is evidence throughout that entire region that the amount of precipitation was formerly much greater than at present, and in so speaking I do not refer to a very remote date geologically, l)ut to a period which was probably post-Tertiary. Indeed, from the present condition of many of the regions in which we know that the early Indians dwelt and which are now perfectly dry, with all sources of water so remote that they can no longer be inhabited, it must l)e inferred that there has been a change in the climate within the period of human occupancy. Certain it is that water is doing very little relatively in this region now, while the agency of wind is conspic- uously marked wherever it can produce elTects. The peculiar form of these buttes is not such as water could have produced, while it is precisely the form that wintl would naturally produce, acting upon the \-ery fine and soft materials, somewhat resembling ashes, that com- pose these Ijuttes. L'4 :mi:s()zc)1c i-j>uka8 of umted states. Further evidence of this is found in (he fuel that in approachino: the general escarpment wliich boiuids these plains tlie l)u(tes tend to lose their isolated character and form ridges projecting out from the cliffs. It never happens that an entire valley or plain is covered by a single system of Inittes. These systems are separated \)\ wide intei'vals, often of nearly fiat country, hut through which it can he easily seen that water once flowed, at least in the form of temporary floods, and in such a manner as to have swept away e\'erv vestige of the foi-mer plateau, and in crossing which there are encountered one or several wide beds to which the term "wash" is popularly applied. In descending the Little Colorado this condition of things, as already remarked, is not met with until within 8 or 10 miles of the Lees Ferry road. A large system of Inittes is then found extending 5 or 6 miles down the river and across the plain to the first terrace, a distance of 3 to 5 miles; then occurs the first wash, 2 miles in width, followed by another system of buttes, which is nearly due east of Tanners Crossing, and in which most of the bones were collected by our party. There is then another wide wash ; the next system of l)uttes, however, does not reach the river, l:)ut trends off in a direction nearly due north. There is still another wash before the great Moencopie Wash is reached, the direction of which is such as to be highly favorable for the preservation of these buttes, and accordingly we find their greatest develop- ment, so far as this region is concerned, along the Moencopie Wash. They do not, however, follow the stream up in the direction of Tuba, but continue to trend northward along the wide valley that lies to the west of Willow Springs and Echo Cliffs. The reason why these conditions are not earliel' met with in the valley of the river is simply that the river does not follow the line of strike, and these beds, being common to the entire member, lie at different dis- tances from the river. Above the point mentioned, therefore, they must be looked for farther in the interior. We found them, in fact, 5 miles east of Black Falls, or 25 miles southeast of Tanners Crossing. The great bend in the river culminating at Winslow keeps these beds constantly so far to the northeast, in a region where it is very difficult to penetrate, that their exact condition for a distance of over oO miles is little known. But farther up the rivei-, where they approach somewhat to the rt^gion of settlement, they again admit of access, and, as already remarked, they TIIK OLDER MESOZOIC OF AKIZOXA. 25 appear in tli'cat foivo in the valley of Lcroiix \\ asli. llci'c tlicv cover an area of nearly 100 square miles and form two ureal amphitheatei's of veritable badlands. The vail, rendei- (hem a magnihcent spectacle. They can he seen from lh(> soudieast for a distance of 20 miles as a white line. Viewed from th(^ top of the mesa otit of which they have been carved, th(> denuda- tion having been ai'res(ed at a particular poini, they reveal more com- pletel}' than at any olhei' place the true character of this memljer. In the Petrified Forest the Leroux Ijeds are also well developed, and the \-ariegated marls ai'e found only half a mile east of the Lower Forest. The buttes her(> are rather large and well tleveloped, and Ijones of the Belodont occui- in them. Li the northern part of the Petrified Forest region the variegated marls lie somewhat farther to the east. What is called the Middle Forest lies in the midst of them, and the petrified wood, as everybody has oljserved, differs here considerably in its constitution and coloration from that of the Upper and Low(>r forests, which lie in the horizon of the conglomerate member. As was remarked when treating of the conglomerates, these variegated marls are actuall>' foiuid stratified between the sandstones by the trans- formation of certain sliales into marls. If these beds are carefully traced a short distance in the direction of the dip they will be seen to thicken very rapidly and soon to take on the character of the true variegated marls. As they start from underneath a bed of sandstone which caps the conglom- erates, and which does not so readily pass into marl, the btittes that are first formed are usually topped out 1 jy a block of this sandstone, and it is necessary to proceed some distance farther in the direction of the clip to reach a point where the sandstones disappeai'. This, however, ultimately takes place and the marl beds thicken to such an extent that they have to be regarded as virtually overlying the conglomerates. In fact, in the bed of the Moencopie Wash, on both sides of which these beds are so well developed, the conglomerates can he seen distinctly passing under the marls. For the purposes of our expedition the variegated marls constituted the most impoi-tant subdivision of the entire group. But as we have seen, their maximum thickness is about 400 feet, and there remain still 26 MESOZOIC FLORAS OF UNITED STATKS. anotlier 400 foot hofoix^ wo I'cncli the base of the painted eUffs. Tliroiigli- oiit the wliole of this fossil wood is abundant, but llae chaructei' of the beds as vai'iefjated marls no longer eontinues. In the lower Coloi-ado \'allev, wluM-e I know it l)est, the variegated marls are succeeded by a sandstone ledge at knist 100 feet hi height, yielding l)lack logs of very fine structm-e. At this point these sandstone beds constitute an escarp- ment and foi-m a small terrace, the summit of which is a dip plane. l'j)ou this lie the remains of the next set of beds, which are somewhat remarkable, primarily in being essentially limestones, but they consist mainly of loose material somewhat resembling dried mortar, for which reason I have designated them mortar beds. They are, however, very irregular in struc- ture and contain much impure flint and large flinty stones. In the midst of them there occurs a true limestone ledge, well stratified, succeeded by a continuation of the mortar Ijeds. In the region mentioned these beds extend to the limit of what I regard as true Shinarump, and petrified wood was found above the limestone ledge. A wider acquaintance with this part of the member shows that the conditions above described do not hold at all points and may even be regarded as exceptional. Nowhere else except at Black FaUs did I find the lower sandstone ledge, and at most other points the limestones gradually supervene upon the variegated marls. In fact, not only the vai'iegated marls, but also the shales of the conglomerate member, which become transformed into marls, are more or less calcareous; and as the entire upper portion of the Shinarump consists mainly of limestones and calcareous materials, we may regard all of tliis, including the variegated marls, as virtually a calcareous deposit. If we were to look for its homo- logue in the Trias of the Old World we shoidd find it in the Muschelkalk, while the conglomerate member might well l)e compared with the Bimter- sandstein, and the Painted Desert formation with the Keuper, to which the French term Marnes Irisees is only locally applicable. In the extensive exposures on Leroux Wash these relations are brought out with great force. Overlying the true variegated marls which stretch out for a distance of 3 miles across the broad eroded valley, the limestone series comes in gradually and scarcely differs, except in the degi'ee of calcareousness, from the underlying beds; but the limestone ledge is ultimately reached and is sharp and definite. It has a thickness of about 10 feet. Over it lie very heavy beds of calcareous materials, TIIK OLDKK MKSOZOIC OF AinZONA. 27 l)eo;innin<; as mortar beds, such as T liavo dcsciilxMl. Init soon takino; on more synnnetrical forms, closf^ly resemhliiisi the mai'l hiitles of the valley helow. The coloi' also changes, and many of the t)utt(>s ar(\ in whole oi' in part , of a dec^p hlne or a hxcly pur])le. These constitute here the hi^i'liest \hh\s (if tlie ShiiK-ii'um]), and fossil wood is al)undaut thi'ou.iihout. Much th(> same coudil ions j)re\'ail in tlie PetriHed I'Orest refiiou, hut tluMlevelop- ment is ]i(>re much less extensive. 77/ A' I'AlXTh'D DESEltr l-OHM ATIOS . It remains to consider the third and highest formation of the Older Mesozoic of Arizona. As already stated, these constitut(> th(» (devated cliffs that hound the valley of the Little Colorado on tiie northeast. Although broken through in man>' places, and practically wanting for long distances, they constitute what may l)e regarded as a great wall, sepai'ating the valley from the region of high mesas that lie in the Moqui and Navajo coimtry. As these l;)eds seem to contain no fossil remains, and as they are throughout the gi'eater {)art of tlieii- extent practically inaccessible l)ecause of the aljsence of water, their detailed study has been neglected, and I was able to actjuaint myscdf with them onl}' imperfectly and at a few points. There is, however, no place where they are better develo})ed than directly east of Tanners Crossing, whei'e we remained longest, and on several occasions the attempt was made to reach them fi'om oui- camp and to examine them closely. Enough was learned to justify the j)osi- tive statement that they consist almost entirely of sandstones, per- fectly stratified, the different layers differing mainly in color, thickness, and fineness of structure. The great central portion constituting the escarpment and having a thickness of about 800 feet is, within these limitations, practically homogeneous. The series l)egins, however, with a bed of orange-red sandstone, highly argillaceous, and soft in structure, easily eroded, and I'eadily yielding to the influence of wind. It has a thickness of about 100 feet, and in the lower Colorado i-egion stretches across the bi'oad valley at the base of the escarpment and lies du-ectly upon the upp(>rmost limestones of the Shinarump. Here it forms pic- turesque and fantastic l)uttes and chimneys standing out upon the plain. It occurs in the same position overlying the Shinarump on Leroux AVash and foi'ming the top of the mesa which ovei'looks the 28 :mi:s()Z()I(' floijas of rNiTKi) statks. amphitheater that I h:ive de.scriljecl. It is also seen above the Shiiia- riimp to the east of the Petrified Forest. It is therefore probal)ly safe to assume that this formation is continuous from Echo ('hffs to the boundary Hn(^ of Xew Mexico. Of tlie painted cliffs there seems to be little more to say. In looking at these cliffs fi'om a tlistance it is seen that they are overlain by a white formation, the nature of which it is important to consider. Before we had visited the region, so as to ()l)tain a close view of them, it was natural to suppose that they might constittite Jin-assic limestones and that the Triassic system might terminate at the line which separates them from the variegated sandstones. But upon close examination this was found not to be the case, and these white rocks were fotmd to consist of sandstones, often very pure and cross-bedded, with scarcely any admixture of marl. These, withovit question, constitute the summit of the Triassic system in this region. The}' are, however, not always white; or at least in some places, as, for example, in the vicinity of Tuba, they are underlain ])y a still thicker bed of soft brown sandstone, which is somewhat argillaceous and easily worn by the wind, forming chimney l)uttes and ruins. This bed has a thickness along the headwaters of the Moencopie Wash of about 200 feet, and is overlain at the highest points by the white sandstones to a thickness of 100 feet more. These sandstones are very porous and all the waters that fall in that region innnediately pass through them; but as they approach the summit of the much harder and firmer beds that constitute the lower formations these waters are arrested and come out in the form of springs, sometimes almost of small rivers, along the crest of the cliffs above the Moencopie Wash. It is on one of these springs that the little Mormon town of Tuba is located, and this is true also of Moa Ave, Willow Springs, and other settlements in that country. Still farther back the Cretaceous lignites and limestones lie vmconformably tipon these tipperniost sand- stones of the Trias, and the Jurassic is wanting altogether. r PALEONTOLOGICAL RELATIONS. Having thus l)riefly sketched the stratigraphical relations of the Older Mesozoic rocks of Arizona, I shall next consider their paleonto- logical relations, in so far as they were ascertained on this expedition, as shedding light upon the age of the group. Tin-: OLDKR MESOZOIC OF AKI/ONA. 29 '/■///•; MDESCOT'IF. FDRMATIiiS. Those beds have jjroved almost eiitiroly barren, no fossil 1)ones haxanji been found in the calcareous marls of the lower part and no fossil wood anywhere in the formation. It is, however, proper to remai-k that in the bed of the Little ( oloiado River about '^ miles bellow Taimers Crossinii", on tlie surfaces of cei'tain fiafjs that underlie the confjlomei'ate mcMnber. and which T had therefore referi-ed to the Moencopie beds, there were found impressions of coniferous twij2;s and short stems, showine, however, very little structure. In some cases the impi-essions showed that the branches sui-rounded the stems in the form of whoris. At first th(\\- recalled impressions of Equisetiun, but fidler investigation showed that they were coniferous stems, with the characteristic whorled bi-ancli- ing of the Araucarian type, to which tlie fossil wood of this region prob- ably also all belongs. As such these \-ague impressions have great value in showing that this type of vegetation continues to l;)e fovmd in the lowest Ijeds in which any plants occur. The qviestion whether these flags actually l)elong to the Moencopie beds is, however, an open one. At the time of our discovery I had no doubt on this point, but after finding the transition beds on Red Butte and in the Little Colorado Valley, it becomes possible to refer some portions of the beds that underlie the true conglomerates in the lower region to the conglomerate meml^er. But it thus becomes scarcely more than a question of names and no longer raises a serious problem. There is only one other reported fact that need be considered in this connection. Mr. P. C. Bicknell, who was my guide in the region of Red Butte, and who has spent much time in studying the rocks of that region, informed me that he once found in the light-colored cal- careous shales at the base of the butte some faint impressions of fern leaves, which wei-e for the most part too frail to be transported, but that there was one which he took to the Anita mine and compared with certain figvn-es in the few Ijooks at hand, and it seemed to him nearest to the figui'es of some species of (!allipteris. The specimen, however, ultimately disintegrated and was lost. On the occasion of our visit he took me to the place and we made diligent search for other specimens, but nothing of the kind could lie found. If the species was really a Callipteris it would indicate a Paleozoic age, but as ]\Ir. Bicknell had read in the l)ooks that treat of Red Butte that it was a Permian 30 .MESOZOIC F]>()UAS OF UNITKI) STATICS. remnant, he n:ituiall\ looked among Permian fossils for a figure with which to compare his specimen and does not seem to have looked fuilliei-. The Trias also yields ferns, and perhaps if hv had examined figvu'es of Triassic ferns he would have found a figure of liis plant. It is a1 least certain that this defective j)iece of evidence is altogether without weight in fixing the age of these beds. THE srilXARrMP FOEMATIOX. So far as concerns the \-ertebrate remains, sufficient has already been said to show that all that were found came from a single phase of the Shinarump formation, viz, the variegated marls. They occur in gen- eral a littl(> higher than the middle and 200 to 300 feet above the top of the conglomerates. Very few other animal remains were found, l)ut Mr. Brown did collect a small number of shells and a few other invertebrates. They are probably for the most part without diagnostic value, but as they have not yet been determined it is impossible to discuss then- significance. The only plant remains that I was able to discover, aside from what l:)elong properly to the fossil trunks, were certain forms occurring in relief on the faces of sandstone rocks and shales. They consist of stems having the Araucarian struc^ture and showing the liranches in whorls, and of the raised casts of small twigs lying across one another in all directions. No signs of the structure nor any carbonaceous material accompanies these impressions, and they seem to have resulted from the etching away of the sandstone from between the twigs while still in the beds, so that when subsequently exposed these markings stand out very^ distincth', though always somewhat worn. They pro1)- ably all belong to the coniferous vegetation, but have very little value in determining its exact nature. For convenience of reference in future I shall name these forms Araucarites shinarumpensis. (See Pis. I, II.) No one who has not visited that I'egion can form an adecjuate con- ception of the inexhaustible quantity of silicified wood that occurs at all horizons. The condition of things in the Petrified Forest has already been set forth by others as well as myself." It strongly attracted the attention "Twcntiotli Ann. Rep. U. S. Geol. Suit., Pt. II, 1900, pp. 324-332. Report on the Potrified Forest of Arizona, by L<>ster F. Ward, Wivsliinjrton, UKM) (spe<-iiil ])ul)li(!i(ion of tlie Depiirtnipnt of the Inlerioi). The Petrified Forest of Arizona, liy Lester F. Ward: Smithsonian Report for ISil!), \Vnshinf.'(on, lilOl , p|). 2S9-307, pi. i-iii (reprint of tlie last with tliree plates added). TllF, OLDKi; MKSOZOIC OF AKIZONA. 31 of all the early exploring part ios that passed through that region, alt lioiigli it is next to certain that none of these parties ev(M' saw what is now called the Petrihed Forest. They all jiassed within a few miles of it, hut either kept in the bed of the Kio I'uerco oi' els(> some distance to the northwest of it. Lieutenant Whipple's pai'ty crossed that sti'eam at Na\ajo S|)i'ings and followed it down at considerat)le distance from the vall(\\- on its I'ight bank, crossing a munhei- of hi'oad washes, which tiiey named. The first of these washes that tlie>- ci'ossed aftei' leaving Xavajo Springs is now called Bonito Creek on nearly all maj)s. It joins the Pvio Pu(M'co about 6 miles below Xavajo Spi'ings. The next wash that the party c]'os.sed they named Carrizo Creek. The third of these valleys or creeJcs was the one in which they found such a great quantity of beautifully colored pet- rified wood, and from this circumstance named it Lithodendron Creek." There are, of course, vast quantities of petrified wood on the slopes of all these streams or valleys. The range of me^sas that skirts the northern flank of the Petrified Forest trends here considerably to the north and reappears on the northwest side of the Pio Puerco only a few miles below Carrizo, to the northeast of which these mesas are worn away much as they are in the Petrified Forest, leaving the petrified wood strewai over the valleys and ridges, so that the conditions obtaining on Carrizo Creek or Lithodendron Creek are very nearh' the same as those of the Petrified Forest. The wood is not so abundant there and is not generally so bril- liantly colored, but some of it is jasperized and is very beautiful. The two great logs that were brought to the National Museum in 1880 or 1881 "In my report on tlie Petrified Forests of Arizona (p. 10), I pointed oiil tlie fact tlmt Lithodendron Creek coidd not by any possibility pass tlirough the present pi^trilied forest, although a numlier of writers have alluded to tlie valley in whicli that forest is located as Lithodendron Creek. And in the Twentieth Annual Report, I't. II, p. 321, I as of logs lie out upon the plain. Tliey have a ivddish-brown color, ai'e very large, and look at a distance like so many rusly locomotive boilers. They are broken aci'oss into sections. Most of the wood at this hoi'izon, however, is not colored, and it has u.sually undergone a higher degree of disintegration than the harder trunks from the conglomerates. It shows the structui'e admirably, at least to all out- ward appearances, and the sections are usually split up into a large num- ber of blocks and ultimately reduced to a mass of c-hips and splinters, which look so natiu'al that they would not be suspected of l)eing petrified unless picked up and examined closely. Many of the smaller l)uttes seem to have l)een occasioned by the presence of logs, which weighted the underlying marls and tended to prevent their being washed or blown away. The result is that many of these buttes have such logs lying on their summits, with the disintegrated material rolling down its slopes. In my report on the Petrified Forests of Arizona (p. 15) I mentioned the statements made by Mollhausen and Marcou that they had seen trunks standing erect and evidently in place, and I quoted (p. 16) Doc- tor Newberry's conclusion, agreeing with mine, that this phenomenon probably did not occur. So far as the conglomerates are concerned, I have seen no reason for altering this conclusion, although I would not be as positive now as I was then that cases of the kind will not be found. But with regard to the trunks entombed in the variegated marls, or next horizon above the conglomerates, we practically demonstrated that erect stumps do occur in them. Within a quarter of a mile of the butte from which Mr. Brown foimd the best preserved vertebrate bones there is a small area, probably 30 or 40 acres, which contains a group of twenty or more such stumps. They are low, rarely rising more than 4 feet above the grovmd, but some of them are large, having a diameter of from 3 to 4 feet. Nearly the entire trunk al)Ove these stumps, as well as all the branches, has wholly disappeared, IjuI the ground is strewn with small chips and bloc'ks. It is a somewhat level HON xi.viji — 0.5 3 34 MESOZOIC FLORAS OF UNITED STATES. afea and the stumps all .stand erect upon it, and many of them are not deeply buried in the earth, but show the natural enlargement toward the roots. A careful examination of these stumps convinced me that thev were not only in place but stood precisely where they grew. Mr. Brown, who discovered this place, is of the same opinion. However difficult it may be to figure to oneself conditions that would preserve trees in an erect position in a sedimentary bed, the fact of their pres- ence in this position seems to Ije conclusive. The most probable theoiy seems to be that, as a matter of fact, the trunks of the trees were not presei'\('d, but only the roots and short stumps, and whatever the con- ditions may have been that were sufficient to preserve large trimks in a horizontal position, the same conditions would surely preserve such short stumps and roots. This seems the proper place to mention another phenomenon which at first was very enigmatic and which can not yet be said to l)e adequately explained. I found on several occasions some peculiar short chalce- donized stems contracted at both ends, many of them broken ti'ans- versely and showing a concentric structure; others split longitudinally. I observed that these varied in length relatively to their diameter until some of them became merely elliptical objects resembling fruits. It was not until the erect stumps, above described, were discovered that the mystery was partially cleared up. Among the chips and blocks that surround these stumps there occur a large number of these ellip- tical fruit-like objects, usually striate on the surface and somewhat flattened, so that the cross section is elliptical. This is the typical form and much resembles a butternut that has lost its exocarp, but a yovy little search reveals the fact that there are great variations from this norm, especially in the matter of lengthening the axis. Then it is soon seen by specimens that can be picked up that the rounded ends represent constrictions between two of the objects, and that they are arranged piimarily along a general axis in a necklace-shaped series. The next and most important fact that comes out is that these rows of nut-like objects adhere to the true fossil wood in the interior of the tnmk and are often actually found in place in the stumps as an integral part of their structure. This, of course, reveals their true character as simply accunuilations or secretions of certain suljstances within the trunks, and everything points to the probability that they consisted THE ()LI)p:it MESOZOIC OK ARIZONA. 35 nripiially of rosin or jiitch. T1h\v may, thorefoiv, poiliaps 1)0 oorrootly designated piteli hlistei's and coinpaivd with the l)hsters of Canada l)alsam that occui' in the hark of the balsam fir. iMU'thei' than this they have no hotaiiical significance. A I'athei' large collection was made, showing all the different aspects and furnishing data foi' the aliov(> conclusion. (See PI. III.) The species may he called Arducdritcs innuilifer, alluding to the necklace-shaped rows of resin dro])s. I had been several times told that petrihed cones had been found in connection with the fossil wood of this region. While at Stanford University in October, 1899, a young man named Dane Coolidge gave me an accoimt of such a discovery made by him and his father at a point 4 miles west of Williams, Ariz., some years before. He said they found large petrified logs, in the vicinity of which they picked up a number of fos.sil cones. He wrote to his father and obtained for me all there was left of theii' collection. It contained nothing recognizable as a cone, l)ut he said that all the good ones liad been given away. As I was going into that country, I thought it worth while to stop and examine the spot, which was ver}- minutely described for me. I found no trunks or petrified cones, l:)ut did find a few pieces of unmistakable fossil wood. The locality is near Supai, on the Santa Fe Pacific Rail- road, where there is a dangerous curve. A short time afterwards I was shown, at the house of Mr. T. W . Brookbank, at Little Spring, on the noi1 Invest side of San Francisco Moimtain, a number of ol)jects which were believed to be fossil cones. They were not sufficient for me to settle the question, but I was told that Mr. Brookbank, who was then away, had much more perfect ones locked up in an adjacent room. These Mrs. Brookbank said were col- lected on blu(^ cla\- knolls near Tanners Crossing of the Little Colorado. The ones I saw were cylindrical bodies, of a reddish-brown color, sur- rounded by quartz crystals, closely inntating the scales of cones. Two weeks later I visited that region, but found nothing that looked like these .specimens. My stay there, however, was too l)rief to enable one to find anything rare, and therefore when I went thei-e this season and devoted more than two weeks to the minute study of the wood-bearing beds of that locality I paid particular attention to the search foi- fossil cones. I found none, but did find many cylindrical ol)jects, some of them surrounded by crystals, which were certainly the same as the 36 MKSOZOIC FLOIJAS OF IMTKD >TATES. alleged petrified cones of Mr. Hiookhank. In passino; his house, on our way to Flagstaff, Mr. Hiowii and I were kindly pennilted l)y Mr. Brookljank to examine all the fossils in his possession. Xone of the supposed cones were sucli. and all of them belonged to the same class of materials that I have described. Although they are not cones, they have some paleol)()tanical interest, and I made a thorough study of their origin and nature. They occur in the Petrified Forest and else- where, are usually spoken of by the inhabitants of the coimtry as stems, and are supposed to be the smaller branches l^elonging to the upper part of the trees which mak(^ up the foi-est. This, however, is an eri'oneous view, and I discovei-ed that they always came out of the interior of the tninks and belong to the bodies of the trees. They vary indefinitely in size and length as well as in texture, and only a few of them are surrounded by crystals. They either have to do with the vascular tissues of the trunks or else they are modifications of the pitch blisters described above, and represent lines along which the resin was disposed to accumulate either during the growth of the tree or, more probably, as a result of the process of mineralization, during which these products were segregated and arranged along certain lines. I lirought with me a sufficient number and variety of these objects to illustrate their true character. The petrified wood of Arizona is found widely scattered over the Paleozoic terrane. One small specimen only, picked up ]\v Mr. P. C. Bicknoll, was found at the foot of Red Butte, which must have come from the conglomerate bed at the summit. But in the vicinity of Williams, both southwest and east of the town, I found many pieces of unmistakable fossil wood lying about among the dark porous rocks of the lava. They all show the effect of heat, are themselves somewhat porous, and have douljtless lost all their minute structure, but tlieir true natui-e as wood can not be dout)ted. I brought away a number of specimens, and also have those collected near Supai two years before. Moreover, I met many persons who reported finding it under similar conditions near FlagstafT and on the north side of Mount Agassiz, as well as farther on in the direction of the Grand Canyon. These occurrences are certainly difficult to explain, especially in view of the fact that, with one exception, no wood has thus far been found below the true Shinarump. It seems necessary to admit that not only the Moencopie Ijeds Ijut also the conglomerate once cov- Tin-: OLDKH MKSO/OIC OF ARIZONA. 37 ered iho entire Pnleozoic terrane, at least as far west as Bill Williams Mountain and Supai. rilK FAlXTh'D DESERT FORM ATI OX . It was remarked that an exception would l)e noted to the general statement that petrified wood, so far as known, is exelusively confined to the Shinarunip. When at Tuba I made an excm'sion to the northeast, over the brown rocks of that i-egion, and in some of the buttcs and chimneys which tliev form I obserxcci black spots. \ casual examination would lead to the supposition that they might be deposits of manganese or linionite. They are mostly l)lack sand, but more extensive observations revealed the fact that they are due to the former presence of tnmks of trees, and in one place I found the remains of a log broken into a number of sections. It consisted, however, wholly of the black sand and had lost all signs of structm-e. Beds of lignite were reported in that general vicinity, and they are prol:)al)ly due to the same cause. SECTIONS. ■ Special attention was paid throughout the expedition to working out geological sections of the beds studied. The more important of them will be introduced here as necessary to complete the cles(;ription of the.se beds. I will begin with the first section made, which resulted from an investigation of the bluffs of the Little Colorado below Tanners Crossing. SEfTlUS I.— CANYON OF THE LITTLE COLORADO. [PI. IV, A-B.] There are a few short canyons in the Little Colorado at various points, l)ut it is not until Tanners Crossing is reached that the canyon becomes continuous to the mouth of the river. For several miles the valley even hei'e is somewhat broad, the bed of the stream usually hugging one bluff or the other; but the bluff's are always 100 to 400 feet high and more or less perpendicular, so that it may be practically regarded as a canyon. The fall of the river is hei-e about 25 feet to the mile, and its course is nearly northwest. As the dip of the rocks is northeast this would practically be the line of strike, but the fall in the rivei- is to be taken into account, and it is also true that just at this point the trend of all the different subdivisions is much more northerly, as I have shown in my discussion of the variegated marl buttes. The consequence is that in reality the bed of the river, from Tanners Crossing on, continues to be lower 38 MESOZOIC FLORAS OF rNirFI) STATES. Limestone Canyon Coconino Point and lower in its horizon, until at last, some 12 miles below, the Paleozoic limestones appear, and in the very narrow canyon suddenly formed at the Ijase of Coconino Point, which forms the eastern _ ro N W escarpment of theColorado Plateau, the limestone rocks rise to a height of lOU feet and constitute the canyon proper. The section along the river between these points was carefully worked out 1 )y me. The most remarkable fact was that about 4 miles above the limestone canyon there occurs a Paleozoic anticline as viewed from the bluffs, the limestones rising to a height of about 90 feet and again descending to the bed of the river, the length of the anticline being about 2 miles. This : of course represents a spur of the Colorado Pla- • teau, nmning out parallel to it in a northeasterly \ direction, which was cut through by the river. = This is shown in the following diagrammatic sec- : tion, 12 miles in length: \ < Description oj the section shown in fig. 1. i Feet. = 1. Carboniferous limestone 100 t 2. Argillaceous shales and sandstones, Mocncopie formation . 200 j .3. Lithodendron member 125 J i. Trap 30 < I Total tbiokness 45.5 - The above are the vertical measurements of f *''^ ^°'""^'" ^'''' the beds as seen in the bluffs. Probably twice that thickness of the Moencopie beds is repre- sented, and only the base of the conglomerate series is here exposed. SECTION II.~MOENCOPIE WASH. [PI. IV, C-D.] This section begins some 5 or 6 miles below the mouth of the Moencopie Wash, opposite the upper end of the limestone canyon. The rocks dip away from the river from the first, and the section remains wholly within the valley of the Moencopie, reaching the bed of it at a distance of lannars Crossing THE OLDER MESOZOTC OF ARIZONA. ^ Limestone Canyon %^Linh Colorado Valley m f'Pi '^ about 7 miles, aiul followinjz; it at that level foi- about (1 miles more, to where there is a decided beml in the stream which comes in fi'om the east. The Moeiieopie h(>r(> runs between high bluff's, and the section shows those of the I'ifjht bank, passing through Tuba, which is 300 feet above the bed of the Moencopie, and continuing on over the l)i'own and white I'ocks. The maximum thickness is ;^,0()() feet and the length of the section 24 miles. The following is the section: Description oj the strttdn shoint in fig. 2. Fpet. 1. ('arliciiiifiM'uus liincstone 1(K) 2. Moencopie fonnation (>()() 3. Litliodeiidron member •'>'*l 4. Variegated marls 4IKI .5. Limestones and mortar beds 2()() (). Variegated sandstones .500 7. Brown sandstones 600 shown in fig. 5. 1. ('ai'l)onifcious sandstone (tliickness unknown). Feet. 2. Mooncopie formation 500 3. Lithodendron member 700 4. Variegated marls -IIX) 5. Limestones and mortar beds 300 6. Calcareous marls 200 7. Orange-red sandstones 100 8. Variegated sandstones 700 9. Brown sandstones 200 Total thickness 3, 100 SECTION VI.— THE PETRIFIED FOREST. Little Coloraao ' /\ [PI. IV, L-M.] I have carried this section as far back on the southwestern slope of the Little Colorado as I was able to find any traces of the saliferous beds. The Carbonif- erous sandstone occupies the northeastern slope of the Pink Cliffs, and the section starts at an elevation above the sea of nearly 6,000 feet. It passes through Woodruff Butte, which is on the bank of the river, and thence on through the Petrified Forest and over the mesa on which the Natural Bridge is situated. I have '// TlIK OLDER MESOZOIC OF ARIZONA. 43 then carried it on a distance of 1() miles farther and l(>t it terminate at the summit of the Shinarump, at an elevation of about 6, ()()() feet ahove the sea. Desert jttinn iij Iht yir/inn shoifit in jitj. fi. 1. Ciulionilcnms siiiirMiintj tlif wt-ll- kiiowu I'aiutfd Cliffs. ' Red-orantre smidstones. Calcareous marls, sometimes wo white, blue, or purple buttes. Mortar beds with impure flintstones. Limestone ledpe, definitely stratified. Sandstone led^re %vith black'logrs; looal. Variet^jited marl>. chiefiy urnilliu-coiis hut somewhat cnlrareotis, often worn into small buttes studdinji the jjlaiiis; white, blue, pink, or red, and beautifully handed, holding bones of belodonts and dinosaurs. CouKlomerates and cross-bedded coarse sandstones. r)fteii with pink and white Striped ehiy lenses, interstratilied with gray aririlliieeoiis shales and \ariei^atcd marls, the latter locally much thickened, tormiiiK brilliantly colored banded cliffs. Dark chocolate-brown arjrillaceous shales, devoid of jrrit and highly charged with 5 salt and trypsum. .Saliferous beds. Dark-brown, soft, argillaceous sandstones. i Arprillaceous, dark-hrown shales, highly sa- liferous and with gypsnm layers, becom- ing: calcareous below. O j_ 0_l_0 ;_0 ' O J_0 ■^^59^1"^ I'jiint.-d De-tt-rt formation. Leronx nieniher. Lithodendron Mocncopie formation. Calcareous shales, mostly white. Saliferous shales as in Nos. 3 and 5 Limestones lerinoidah conformable. )r sandstones. Vn- FiG. 8. — Geological column of the Older Mesozoic of Arizona. 46 MESOZOIC FLORAS OF UNITKI) STATES. tlie drainage and all the principal localities are shown and the lines of the sections indicated as follows: A-B, Section I: Canyon of tho little Colorado. C-D, Section II: The Moeiuopie Wash. E-F, Section 111: The lowei- Little Colorado Valley. G-H, Section IV: Winslow. I-K, Section V: The Leroux Wash. L-M, Section VI: The Petrified Forest . ' N-0, Section VII: Red Butte. P-Q, Approximate line of strike. P^RT T r. THE JURASSIC FLORA (CoxTixtED). rilK JITKASSIC FLORA OF OltKlJOX. A sufficiently full account of the expedition made in 1899 to the Buck Mountain region of Oregon, as also of previous studies of and col- lections made in that region, was given in the first paper." As stated there, all the specimens that had ever been sent to Washington, including our large collection of that season, and the collections made by Mr. Storrs during previous years, were sent to Professor T'ontaine foi' elaboration, and their study had been begun by him before tliat paper went to press. Buck Peak It was completed in tlie spring of 19U1, and Professor Fontaine's report, including descriptions of the species and careful indications relative to the illustrations, was sent on in April. The types have now all been figured, a new process (the Williams process) having been employed. It therefore only remained foi' me to prepare the report for publication. The following paper is essentially the report of Professor F'ontaine, only, as in former cases, it has been necessary for me to prepare the synonymy of the previously described species and attend to the systematic arrange- ment. This, therefore, and a few footnotes, to which my initials are attached, are the only parts for which I am responsible. The accom- panying sketch map of the region and section, prepared by Mr. J. S. Diller, will make the geographical and stratigrajihical relations clear. (See PI. V.) "Twcntii^tli Ann. Hep. U. S. Gool. S>irvc.v. I'l. H. lOIKI, p|). :-!fi.'S-.'^77. 47 48 MESOZOK' FLORAS OF LMIFD STATES. THE JURASSIC FLORA OF DOUGLAS COUNTY, OREG Bv Wii.i.i.vM M. Fontaine. The plants descri])0(l in this paper wei'e obtained from strata that occur in Douglas County, Oreg.. on or in the vicinity of Buck Mountain. Some fossils were originally discovered Ijy Mr. Aui'elius Todd, in 1872, at a localit>' about 300 feet l^elow the summit of Buck Mountain, which has subsequently been named Todds Gulch. Since that time additional collections have been made both at that locality and at others discovered since.the original find of Mr. Todd. The new localities occur north of the mountain on its slopes, and south of it in the vicinity of Nichols station. Professor Ward has given an account of the discovery of the different localities, and also of the stratigraphical investigations made by Mr. Diller and Mr. Brown, at the time that the last and largest collections were made by himself, aided by Mr. Storrs. This account will be foimd in the first paper." From the facts given in that paper and in a letter from Professor Wanl, and also from a few notes kindly furnished l)y Mr. Diller, the details that follow, regarding the position of the plant localities and their geological relations to one another, are obtained. The study of the geolog- ical relations of the strata containing the plants was carried far enough to show that, from the most northern to the most southern localities showing plants, they are continuous and form one group. At remote points, however, the exact stratigraphic relations of the plant-bearing beds were not made out, so that it can not be stated whether or not they are the same. Buck Mountain is about 8 miles due west of the town of Riddles. It is about 3,500 feet above sea level, and rises 2,000 feet above the streams that flow along its base. On its west side, flowing north, is Olalla Creek. A ])i'anch of this creek flows westward along the northern base of the mountain, and is locally called Thompson Creek, but on the Land Office map it is named Hunter Creek. The plant-bearing strata are exposed on Thompson Creek, and at several spots collections were made from them. The strata on the creek tlip toward the west at an angle of from 35° to 40°, and have a strike of N. 15° to 20° E. In passing south, however, into "TweiUictli Ann. Kcp. U. S. Geol. Survey, Pt. II, ISHK), pp. ;«iS-;57 JURASSIC FLORA OF DOUGLAS COUNTY, ORECl. 49 Buck Mountain, tho strike varies considerably. It is not stated how much of the mountain is composed of these Ix'ds, hut the higliest of them occur about 300 feet l)elow its top, where they are overlain by a heavy conglomerate of Lower Cretaceous age. Here the fii'st discovery- of the fossils was made by Mr. Todd. Plants were found in the strata in Buck Mountain at least 30 feet below Todd's original locality. On Thompson Creek the plant beds are flanked to the east by a thick mass of sandstones and conglomerates of unknown age, which dip under them. Still further east this last-named group is bounded by a belt of igneous rocks, to the east of which lies a l)elt of sandstones which contains invertebrate fossils of Lower Cretaceous age. These sandstones dip westward, as if lying imder the plant l)eds. They may be dropped in this position l)y a dislocation. The sandstones are bounded on the east by a great mass of serpentine. At the western end of the section, on Thompson Creek, the Lower Cretaceous conglomerate, which overlies the plant beds in the top of Buck Mountain, is absent, it having been removed by erosion before the deposition of the Eocene. This last immediately overlies the plant beds. Professor Ward states that on Thompson Creek, the first of the group now in cjuestion that was found to contain plant fossils, is a slate that lies to the west of the sandstone and conglomerate mass of unknown age above mentioned. This occurs nearly due north of Buck Mountain. This is stratigraphically the lowest plant bed on the creek. The stratum with plants is only a few feet thick. This for distinction I shall call plant bed No. L It is overlain by conglomerates 50 feet or more in thickness. The conglomerate has overlying it another bed of slate similar in general appearance to the first. This also contains plants and yielded much the larger part of the specimens collected there. It ma}' l)e called plant bed No. 2. In the vertical section it is about 75 feet above bed No. L This seems to be the highest Ijed geologically from which collections were made on Thompson Creek. The upper slate is overlain a short distance to the west by Eocene beds. Mr. Diller and Mr. Brown followed the group containing the plants southward into Buck Mountain, proving the identity of the plant beds of that moimtain with those on Thompson Creek. From the moinitain the strata were followed southward to the vicinity of Nichols station, where .MON XLVIII — 05 4 50 MESOZOIC FLORAS OF UNITED STATES. the mot;t southerly collections were made. Nichols station is on the Southern Pacific Raih'oad, 7 miles due south of the Thompson Creek plant beds, and the plant localities near this station lie in a north-south line that passes through those on Buck Mountain and ends with the localities on Thompson Creek. The geology of the region near Nichols station is more complicated than that of Thompson Creek, and, as will be seen from the statement that follows, there would seem to be some change in the rock character. Mr. Will Q. Brown first discovered plants in the railroad cut near the whistling post for the station, and made a small collection. Pro- fessor Ward says, in speaking of his visit to this locality, that " very little additional to ]\Ir Brown's collection was found in the railroad cutting." It was seen however, that the same slates occur here as in the Buck Mountain district and that they came from the north in a regular way. At this point the course of Cow Creek is nearly along the strike of the slates, and they cross the creek very obliquely and follow the bed of the stream for some distance, giving an exposure that permitted a large collection to be made. The points of difference from the exposvu'es farther north are these: The slates near Nichols station are nearly vertical and have a high dip to the east instead of dipping westerly, as on Thompson Creek. They have a thickness of about 200 feet and show no conglomerate bands. As there had been no continuous tracing of the strata from Buck Mountain to this locality, the stratigraphical work does not show whethei' oi- not the plant beds here are identical with those on Thompson Creek. From the preceding statements it will be seen that the plants described in this paper come from three regions separated by intervals in which no plants were collected. The Thompson Creek region is separated from the Buck ^Mountain localities by a comparatively short interval, while the Nichols station localities are much more remote. The collections made at the different spots, at different times and by different persons, are of very unequal value, for they range from only one or two specimens in some cases to hundreds in others. This fact should be borne in mind in noting the distribution of the plants. The absence of a given plant from a particular locality may mean, not that it was really aljsent, but that the collector failed td obtain it in his imperfect search. In order that some idea may be had of the great difference in the size of these collections I will give a brief account of them, stating the localities from which they JrUASSIC FLOIJA OF DOIdFAS COINTY. OKFJi. 51 were ohlnined. For convcincnco of I'cfei'cnce the localities will \>v miin- l)ered and the reference of the different fossils to them will he liy these iiumhers. In the case of some collections the specimens were connted. In other cases the nnmher must he indicated in some other way and estiniat(Ml. The followinfi ai'e the localities: Liiciilih/ \(i. I. 'I'liis is 'Pctdd's lirsl discov crcd lociility. It is (in tlic oast side (if Buck .Moiiiiluiii, HOO feet liclow llic l(i|). ( )!' M|-. 'I'odd's cdllccl iiitis (inly .'? siM'ciincns ai'c anidni:; those axMilalilc lor cxaniinatidii. Fidiii lliis i(icalil\ M]'. Storrs sulisi'((iiciU h (•(illcctcd, in S('|)t(Mnl)('r, IS'JT, alxiut 50 spcn'imcns, and >lill later Pnil'i'ssdr Ward and Mr. Storrs obtained 'IT^ specimens. Locdliti/ .\'(). .!.- This locality occurs on Tlionipson Ci-eek, at llie ndilhern foot of Buck Mduntain, near tlie dam of the Day Hydraulic (iolil Minin<; Cdiiipaiu . From this locality Mr. Storrs, in the fall of IStlfi, obtained a small collection of about 30 specimens, and on June 30, 1897, he made here a somewhat larjijer cdllectidn of about 50 specimens. In September, 1897, Mr. Storrs obtained from the same locality aliout 70 specimens. These plants come from th(> lower slate bed. lied No. 1, mentioned above. Some of the followinf^ collections, made later by Pro- fes,sor "Ward and Mr. Storrs in this region, may have been obtained from the spots at which Mr. Storrs had jireviou.sly collected. Locdlity JVo. 3. — This is in a ravine at the foot of Buck Mountain on the I'iijht bank of Thompson Creek immediately below the Day hydraidic dam. Only 1 sjiecimen was collected here by Mr. Storrs, on September 12, 1899. LocalUy No. J^. — This is 200 yards farther down, or to the southcnist of locality No. 3, in a ledge on the right bank of the same stream. Here, on Septendier 14, 1899, Professor Ward collected nearly 100 specimens. Locality No. 3. — This is on the left bank of Thompson Creek, on the \\('stern slojie of a ravine coming from the south and 300 yards below the Day hydraulic dam. Professor Ward collected here, on September 12, 1899, 15 specimens. Ldcnllty No. 6'. — This is 100 yards southwest of locality No. 5, in a low ledge on the left bank of Thompson Creek. Professor Ward collected here, on Sc]itcmlier 14, 1S99, about 40 specimens and small fragments. Locality No. 7. — This occurs in tlie bed of Thompson Creek, in the right bank, beginning nearly opposite to locality No. 6, but extending some 50 feet along the stream. Professor Ward and Mr. Storrs collected here, vSeptember 12-14, 1899, several hundred fine specimens. This is decidedly the largest, collection made. It much surpasses the other collections in the size of the sjx'cimens and in the pci'fccddn of ihc plants. The plants collected here and at locality No. 6, come from the slates above the conglomerate. Locality No. 8. — This occurs 200 jard-s east of locality No. 2. Mr. Storrs collected here, in September, 1897, about 10 specimens." "He seems to have exhausted the locahty.as no plants could lie U:\un\ there, even by him, in 1899. He said that liis specimens came from one bowlder which he broke up. — L. !•". W. 52 MESOZOIC FLORAS OF UNITED STATES. Locality No. 9. — This is 200 yards nortliwcst <>1' locality No. 1. Diiicr and Brown collected here, on Sc'pti'iiiIxT l.'i, ISOit. only 1 specimen. Localitii Xo. 10. — This is situated one-third ol' a mile northwest of locality No. 1. Diller and Brown collected here, on September 13, 1S9'.), only 1 specimen. Locality .Vo. //. — This is a gidch called by the collectors "Ginkgo Gulch." It is on the sl()|)e of Buck Mountam, northwest of locality No. 1. Diller and Brown collected here, on Septeml)er 1.'?, 1S9'J, 2 specimens (comiterparts), with impres- sions of a Ginkgo. Locality X<>. 12. — This is on the main spur running north from Buck Peak, in strata (Jurassic) close to the Lower Cretaceous. Mr. Diller collected here, on September 14, 1S99, 5 small specimens. Locality No. L"i. — This is in the next gulch north of Todds Gulch, at about the same altitude as locality No. 1, and it is perhaps the same locality as No. 9. Mr. Storrs collected here, on vSeptember 1.5, 1899, 5 specimens and several fragments. Locality No. 14- — This is on Thomp.son Greek, one-fourth of a mile above locality No. 2. Mr. Storrs eollected here, on Octoiier 2.5, 1S97, al)out 25 specimens. The rock here is much like that of locality No. 7, and splits well. Locality No. 15. — This is on Seven Spring Ridge, a mile east of Buck Peak, and one-half mile east-southeast of locality No. 1. Professor Ward and Mr. Storrs collected here, on September 15, 1S99, 13 specimens. Locality Na. in. — This is in Todds Gulch, 20 feet below locality No. 1. Pro- fessor Ward and Mr. Storrs collected here, on September 15, 1S99, 14 specimens. Locality No. 1 7. — This is in Todds Gulch, 30 feet below locality No. 1 . Professor Ward and Mr. Storrs collected here, on September 15, 1899, over 100 specimens. Locality No. IS. — This is in a railroad cut near the whistling post, half a mile north of Nichols station. Here Mr. Will Q. Brown collected about 30 specimens, and Mr. Claude Rice obtained, at another time, 2 specimens. Later, on Septemljer 17, 1899, Professor Ward and Mr. Storrs collected at this spot over 100 small specimens. Locality No. 1.9. — This is in the bed of Cow Creek, on the right bank, one-half to three-fourths of a mile north of Nichols station. The locality contains the con- tinuation of the slates that yielded the plants in the railroad cutting. They are here nmch better exposed. From these slates Professor Ward and Mr. Diller, and Messrs. Storrs and Brown collected, on September 18 and 19, 1899, several hundred specimens. This collection is not so large or so fine as that made at locality No. 7, but mucli surpasses any of the others. Locality No. '20. — This is in the bed of Thompson Creek, a little east of north of Buck Peak, on Josten's ranch, at the spot where parties camp. This is quite unimportant, as Mr. Brown collected here, in 1898, 2 specimens only, showing faint traces of a plant. From this account of the localities and the collections made at them it will be seen that localities Nos. 3, 9, 10, 11, and 20 affoixl such small collections that they are unimportant. They show nothing that indicates JIHASSIC FLORA OF DOrCiLAS COUNTY, ORECx. 5o an age different from tluit of the loealilies from wliicli larger collections were made. In estimating the fitness of tlu'S(> collections to give an i(l(>a of the flora of the time in which the slates were deposited, we must bear in mind that we can not judge from the mere innnlier of specimens. The later collections, which are by far the largest, were made under Professor Ward's supervision, ;ind in part by himself. From this cause the collections con- tain a much larger proportion of specimens showing different plants and significant parts of plants than they would contain if made 1)>- one unac- quainted with fossil botan>-. In the latter case a large percentage of the specimens are duplicates that throw no additional light on the character of the plant or else are very vague impressions that can not be determined. DESCKIITIONS OF THE SPECIES. Pliy It nil 13J^ YOPH^^T^. Class HEPATIC.E. Order MARCHANTIALES. Family MARCHANTIACEiE. Genus MARCHANTITES Brongniart. Mabchantites erectus (Bean) Seward?'' PI. VI, Figs. 1,2. 1864. Fucoides erectus Bean in Li'ckent)y: Quar. Journ. (leol. Soc. London, Vol. XX, p. 81, pi. xi, figs. ;ia, 3b (erroneously numbered 2a, 2b on the plate). 1S69. Tlalheris erectus (Bean) Schimp.: Pal. Veg., Vol. I, p. 185. 1898. Marclauitites erectus (Bean) Sew.: Fossil Plants for Students of Botany and Geology, p. 2.33, fig. 49 on p. 233. «I sliall follow, as nearly as practicable, in this paper the system of Adolph Engler, as contained in Die natiirlichen Pflanzenfiiinilion of Engler and Prantl, continued by Engler since tlie death of Prantl, and perfected in the latest edition of his Syllabus. The names of the several gi'oups, however, will not be in all cases those of Engler, but will confonn to the new Code of Botanical Nomenclature adopted by American botanists and published in May, 1904. In my first paper the Bryophyta, Pteridophyta, and S|)prniatopliyta were called subkingdoms of the vegetable kingdom in general. The American code proposes the lerrn ' pliy- lum" for these, conforming to zoological usage. — L. F. W. ''Mr. Seward, in his .Jurassic Flora of the Yorkshire Coast, p. 49, includes in the synonymy of this species, without questioning them, the Fncoides arcuaius of Ijindley and Hutton, published in 1837, and the Sp}i:'n>- coccites arcualiis, which was the name given to tliis form by Presl in 1S3,S, and takes up for a specifi<- name the Fucoides erectus of Bean, figured by Lcckenby in LStil. If the Fucoides tiri-nalu.s is the same as tlie 54 MKSOZOIC FLORAS OF UNITED STATES. Leekenby lias (lescril)O(l from the Bcarlwroujili Oolites, as Fucoides eredits," a sinjjiilai' plant that resembles one found in a single specimen at locality Xo. IS. Owintj; to the small amouni of material, this can not be certainly identified with Leckenby's fossil. Seward has given Leck- enby's plant the name Mdrclutntites erevlusj' The Oregon fossil shows only the imprint, no plant mutter l)eing preserved. It is composed of a rather flexnous stem, apparently once cylindrical in form, that sends off oblifjuely and sparingly shoi't stout Ijranches that have obtuse ends. The branches maintain their width to their ends and are nearly as strong as the axis from which the>' are sent off. In the main stem, if it can be called such, and in each l)ranch, there is a single flexuous nerve cjuite dis- tinctly shown. (Jn the stem and branches there is a vague reticulation on each side of the midnerve, which appears to be caused by depressed areas. In tlie center of the depressed areas there is apparently a small prominence, possibly due to a sorus. Leekenby describes his plant as having a midnerve in each branch, on each side of which there is a fructi- fication composed of one or more rows of ovate ^•esicles immersed in the frond. The mode of branching of the Oregon fossil differs from that of Leekenby in l)eing not so palmate. It is similar to that of Brachyphyllum and the plant may be realh' a twig of that conifer. P»liyliiiii FTERIDOfHYT^ (P^erns and Fern Allies)/ Order FILICALES. Ferns. — Fems arc not rare at some of the localities and they show a decided difference in distribution, for in some places they are almost entirely wanting, being most deficient where the cycad remains are most F. erectus, the combination should have, by the rules of nomenclature, the earlier specific name. In his dis- cussion, however, on the next page, after examining both the types, he says that "the specimen to which Lindley and Hutton applied the latter name was much more imperfect than Leckenby's type, and it is-not certain, though highly probable, that the two are specifically identical." I have not thought best, therefore, to change the combination, but the only logical way to escape from the didiculty is to omit the doubtful name entirely from the synonymy, whii^h I have done. — L. F. W. "On the sandstones and .shales of the Oolites of Scarborough, etc.. by .Tohii Tji'iUi'ti1>v; Quart. .Tourn. Geol. Soc. London, Vol. XX, 18(54, p. 31, pi. xi, figs. Sa, 3b. '' In his Yorkshire Klora. HKK), he rede.scribes the species on p. 49 and reproduces on p. .")0, lig. 2, the ligiire cited from his Kossil I'lants in the above synonymy, which is from Bean's type specimen in the Woodwardian Museum represented by I^'ckenby in his fig. 3a. He finds, however, in the British Mu.seum of Natural His- tory" at South Kensington another specimen (No. V. 3652) which he figures on pi. xix, fig. 2, of his Yorkshire Flora (seep. 51).— L. F. W. ''See footnote to Br>-ophyta, p. .53. JIHASSIC KI^OHA OF DOlCiLAS COl'NTY. OHEG. 55 •abundant. The specimens are generally very fragnieiitaiy, aiul, what is noteworthy, the parts preserved are often in a good state of preservation in that the epidermal tissue is intact, and the {)laiit substance gives a black carbonaceous film on the rock. The parts do not seem to have suffered much from maceration due to long floating in water, hence the fragmentary state nnist be produced by some other cau.se. The coniferous fo.ssils also show a great conmiiiuition of parts, with a good preservation of tlie })laiit sul)stance. The cycads do not sliow so extensive a laceration, although they, too, ;iie niucli broken, while the parts that are shown are wonderfully well preserved. Family CYATHP]ACK.E." Genus DICKSONIA l>Il.-iiti('r. I)u KsoNiA oREGONENSis Foiitaiiie n. sp. PI. VI, Fijis. 3-9; PI. VII. 1898. l>ni' large. Many of the specimens present the lower surface of the foliage uppermost, so that the soi'i are well shown, and in many cases they are remarkably well preserved. They are placed within the margin and are attached to the end of one of the lateral nerves. In the case of the entire pinnules there is only one sorus, and this is on the end of the lateral nerve given off from the base of the midnerve on its upper side. The sorus is so large that it covers most of the surface of the anterior base of the pinnule. The pinnules with dentate or lobed mar- gins have a sorus for each tooth or lobe placed on the ends of lateral nerves and covering the anterior portion of the tooth or lobe. Often nothing but these large sori can l)e seen in such pinnules, and by their position they outline the form of the pinnule. The sori are covered by a thick durable indusium, which is appar- ently inserted under them. The sori stand out strongly convex and leave deep pits in the rock. They are globular in form, with a slight elongation. Owing to distortion they vary a good deal in their present form. When, however, the specimens present the upper surface upper- most, so that the sori must be seen through the lamina, they appear much smaller and are punctiform elevations. The fructification as made out is near enough to that of Dicksonia to justify placing the plant in that gi-oup. .Hh'ASSIC FLORA OF DOUGLAS COUNTY, OKEG. 57 To judfic fi'oiii tlic spociinons, the plant sliow.s a ina-rked tendency to fiTietification, for most of theni are fertile parts. It is prohahh' a new speeies. The stei'ile eiitii(> |)innul(>s r(>seinl)]e thos(> of Heer's Dick- souid (//Y/r/Y/.s'," from the .iui-assic of Asia, hut the ultimate pinna' ai'e shorter, and Heer's {)lant lacks the heteiomorphous featui'e in the pinnules. Heel- calls attention to the resemhlance of iiis plant to the Sclc- 7-njiteris Poniclii of Saporta, from the .Jurassic of France. This resem- blance exists, hut it is not sufficient to justify identifying the Oregon foi'in with Sapoi'ta's fossil. Xo fructihcation exists on either of these previously de.scrihed fossils, and in view of the predominance of it in the Oregon fossils it is not likely that it would be wanting in the for- mer if they are identical with the latter. In sliape and size the sori of l)ii'l:>t()ni(i (irrgnne7isiff agree pretty well with those of Heer's Dick- sonia clavi/ics,'' from the Jurassic of Siberia, but the fertile pimuiles are not, as in that plant, contracted to stalks. So many specimens of the plant now in question were oljtained that a pretty full repre- sentation of it may be given l)y selecting parts from different positions on the compound pinnae This is necessary because of the small size of the fragments that are preserved. PL VI, Fig. 3, represents a portion of a penultimate pinna with several attached ultimate pinnae, as well as several unattached ones, that apparently were once attached. This is the only specimen in hand that is credited to Mr. Todd's collections. It presents the upper sur- face of the plant uppermost and shows the sori as they appear in such a case. When this specimen was the only one available, I was let! to regard it as Asjyidium monocarpuni, a fern found in the Lower Creta- ceous of Great Falls, Mont. This is the specimen referred to as Dryoptens mouncarpa by Pi'ofessor Ward in the paper quoted above (]). 8()9). Specimens showing the same character are not unconmion in the collections made since that of Mr. Todd, and they show that the plant is Dicksonia oregonensis. PI. VI, Fig. 4, shows a single pinnule " Flor. Foss. Arct., Vol. IV, I'l. U (Beitriige ziir .liira-Klora Ostsibiticns mid des Amurlandes), p. 92, pi. xvii, fig. .3. ''Op. cit., pp. ;«-34, pi. ii, Cg. 7. I'The genus name Dryopteris (.Vdanson, 1763) liii.s priority over .Vspidiiun (Swartz, 1800) by twenty- seven years. Dr. Knowlton referred Professor Fontaine's speeies (inoiiomrpa) to the former genus in his Catalogue of the Cretaeeous and Tertiary Plants of North Ameriea (Bull. U. S. Geol. Survey Xo. l.')2. 1898), p. 92.-L. F. W. 58 MKSOZOIC 1 LOIJAS OF rMTKD STATKS. enlai'ftcMl. l-iii. 5 gives Ici-iiiiiial portions of two ])enultimatp pinute (•arryiii<: slorile piniuile\s, with entire marfiins (entire pinnules), and henee they come prohal)ly from higli up on the compound pinna. PI. VI, Figs. and 7, represent magnified pinnules to show details. PI. \'I, Fig. S, i-epresents portions of pemdtimate pinna\ with fertile entire and dentate pinnules, that prohahly nearly correspond in position with the sterile parts given in Fig. o. PI. \1, Fig. 9, shows an enlarged pin- nule of Fig. . 175 [33], pi. ix, fig, 12: pi. xii, figs. 7-16; pi. xiv, fig. 17. 1900. Dicksonites clavipes (Fleer) Sew.: Manchester Memoiis, WA. XI.1IV, No. 8, P-6. 1900. Thyrsopteris Maakiana Fleer ? Font.: Twentieth Ann. Rep. U. S. Geol. Surv., 1898-99, Pt. II, p. 343, pi. xlix, fig. 1. 60 MESOZOK^ FLOKAS OF UNITED STATES. IflOO. Coniopieris hymenophylloule.i (Brongn.) St'\\. : .lur. I-'l. '^'orksli. Coast, p. 98, pi. xvi, figs. 4 fi: 1)1. xvii. (igs. 3, 6-S; pi. xx. ligs. 1. 2: ])\. xxi. figs. 1-3, 3a. 4. 4a. .\ small Sphoiioploris-liko f(Mn, with fructification, was found at locality Xo. 19. In the shap(> of it.s pinnules it much resembles .some of the sphenopterid forms from t he Lower Oolite of Yorkshire. Seward," in his paper on the plants of the Manchester Museum, pives good rea- sons foi' adopting; the name Conioptei'is for some of the sphenopterid forms of the Yorkshire Lower Oolite, ;uid unites a numl)er of them with his species Conioptrris hyvwnojyhylloidcs. In his recent work on the Yorkshire Fossil Plants he gives a number of figures of this species. Some of these agree so well with the Oregon fossil now in question that I have no doubt that ii is the same species with the English one. As, however, the amount of niaterial is so small. I do not positively identify it as such. The Oregon fossil occiu's in only one specimen in counterparts. This shows a small fragment of a penultimate pinna, having several ultimate ones that are neai'ly entire. The ])lant is l)eautifully pre- served. The leaf substance is very thick and it appears wrinkled by the strong nerves and the sori. The ultimate pinnae are very short, the longest being hardly 15 nun. long, while their entire expanse at ba.se, the widest part, is only about 7 mm. The pinnules are minute, the largest basal ones being not more than about 4 mm. long and a little over 'A nun. wide. They vary a good deal in shape as well as in size, according to their position on the ultimate pinna. The rachis has a narrow wing from which the pinnules rise. On the upper side of the rachis the piniuiles are larger than on the lower side. The basal pinnule on the upper side is larger than the others on that side. This larger pinnule is subquadrate in form. The others on the same rachis become more rounded, elliptic^al, and even club-shajxMl toward the ends of the pinna*. The pinnules on the lower side aic all ellij)tical, passing to rounded and club-shaped forms. The sulxjuadrate, and indeed all the pinnules, are attached by much narrowed bases. The lateral nerves in the larger pinnules are pinnately, but very obliquely, placed on the midrib. In the others there is a parent nerve. o Notes on some Jura.ssic plants in the Manchester Museum: Manchester Memoirs, Vol. XLIV, No. 8, 1900, pp. 5-8. .in; Assic iLOKA OK norcrLAS corNTV. oinx;. (>] PI. \'lll. liir. 1. shows l)()tli coiiiiKM'pai'ts natural size. PI. \'ni, I'ijr. '_'. ^i\('s, (Milarjieil, an u|)|)(m- hasal [)iiuiiil(\ and l''ig. '.i one from the lerniinal poi'tion of tlie pinna. (Ipiuis TlIYKS()l''li:i;i> Kunl/.o." 'riiVKsoriKHis MruHAVANA (Broiifiniart) Ileor.* IM. \ 111. Fi.<;s. III. ISIili. Picnptf ris M iirnn/diKi Broni^n.: 1 list . \'i'<;. Foss., p. .S.5S, pi. cxxvi, fi<;s. 1, lA, 2-4, 4A, ry. r>A. "Mr. .Sewniil very iiatmally iIdiiIus ihc iicciMicnci' in a fossil state of a inonotvpir jjenus of ferns now living, hut confined to the island of .hian Fernandez, and he thinks that the Cretaeeons species helong to the extinct genus Onychiopsis of YoUohaina and the .Iura.ssic ones to lirongniart's C'oniopteris. In the Fif- t(M'nth Annual Ueport of the United States (ieolosical Survey, I.S!)."), pp. ;j,S.'{ 381, 1 discussed this question, referring to Thi/rsopteri': Miirraijuiia (Brongn.) lleer from the Oolite of Yorkshire, saying: " Hrongniarl had already pointed out the re.setnhlaiice of his I'ecopteris M iirmii' forked at their tips or else simple. In the serrately dentate and lobed pinnules the lateral nerves, one for each incision, are once forked, the forking occurring more deeply as the incising is deeper. The simple lateral nerves and the parts below the forking in the forked ones are parallel to one another. All the nerves are remote and distinct. The pinnules, in becoming incised, change their form, being oblong, with little or no falcation. 64 TklKSOZOIC FLORAS OF UNITED STATES. They are now slightly {'ontracted at the attachment of their bases, but are still decui'rent to form a wing. When the incision of the mar- gins is least, they are undulate or crenate. With deeper incision they are serrately dentate or lobed, the teeth and l()l)e8 l)eing a(;ute and having more or less of the character of the simple pinnules into which the lobes, with increasing deptli of incision, finally pass. These incised pinnules are attached at an angle of about 45°. The fertile pinnules show a greater tendency to obtuseness, the teeth and lobes being often less sharply serrate. The sori are placed much as they are in Dkksonia oregonensis. In the simple pinna? there is only one sorus, which is placed within the margin, on the end of the unforked branch of the nerves, in the anterior basal part of the pinnule. The great size of the sorus causes it to fill much of the basal part of the pinnule on tlie upper side of the parent nerve. In the incised pinnules there is one sorus for each tooth or lobe, and they are placed on the ends of the lateral nerves. The sori are without indusium, and under the pressure to which they have Ijeen subjected appear as flat, rounded patches, with a granulation, which is apparently due to the naked capsules. The granules are grouped around a central point, which seems to have been the receptacle. When, however, the sori are seen with the upper surface of the pinnules pre- sented uppermost, they are a good deal disguised and their true natiu'e could not be made out. They are then smaller and appear as rounded prominences. The fructification seems near enough to that of Poly- podium to justify the placing of the plant in that genus. Owing to the considerable number of specimens that were obtained, a good idea can be formed of the plant. It must have been of large size, with widespread pinnae, for in their great length they contrast strongly with those of Dicksonia oregonensis. As in the case of this last-named plant, by putting together fragments from different parts of tlie compound pinna a partial restoration may be made. It is a much more elegant fern than D. oregonensis. PI. VIII, Fig. 12, gives fragments of two penultimate pinnse, which carry ultimate pinna? and sterile simple pinnules. The penultimate pinnae were evidently once united to a pinna of superior order. The fragments come from near the ends of the penultimate pinna? or else from high up on JTRASSIC FLORA OF DOUGLAS COrXTV. OREG. (to a (.■umj)t)Uiul pinna. 1*1. \'I11, l"i fertile piimules differ somewhat from sterile ones that eome fi'om similai' parts of the compouiul pinna. They are wider in proportion to length and more obtu.se. PI. IX, Fig. 1, rejtresents a larger specimen, having sterile piiuudes, from a position lowei' down on the compound pinna than that shown in PL VIII, Fig. 12. I'liey aic undulate or dentate. PI. IX, Fig. 2, shows an enlarged pinna. PI. IX, Fig. ."!, I'epresents a frag- ment from a still lower position, carrying portions of ultimate pinnae and sterile, serrately toothed pinnules. PI. IX, F'igs. 4 and 5, show magnified pinnules of this. PI. IX, Fig. 6, gives terminal portions of ultimate pinnsp that apparently correspond with the j)ai'ts represented in Fig. 3, but these are narrowed by distortion from pressin-e, the margins being pressed into the rock. If we suppose the parts represented in Figs. 3 and 6 to be com- bined, we may get some idea of the expanse of this elegant fern. PL IX, Fig. 7, gives a small fragment comparable to the portion shown in Fig. 3, l)ut from a lower position on the compound piinia, where the pinnules, by deeper incision, Ijegin to pass into ultimate pinnte. PL IX, Fig. 8, shows one of these pinmdes enlarged. PL X, F^ig. 1, depicts a specimen of con- siderable size, it being a portion of a peiudtimate pinna to which are attached a consideral)le number of ultimate pinna^, carrying crenately toothed fertile pinnules. Unfortunately most of these pinnules are con- sideral)ly distorted, being narrowed by having their margins pressed down into the rock. PL X, Fig. 2, represents the terminal parts of several ulti- mate pinna' with fertile pinnules corresponding in position to the fragment represented in Fig. 1, l)ut in this case they are not distorted. From these the true form of those given in Fig. 1 may be determined. PL X, Fig. 3, shows a single pinnule of this specimen enlarged to show the sori. PL X, Fig. 4, shows a fine specimen. It is a fragment of a penultimate pinna with ultimate pinnae, carrying fertile pinnules of the largest size. They are more deeply incised than the pimiides shown in Fig. 1, and apparently come from lower down on the compound pinna, where the piimules tend to pass into ultimate pinna\ These fertile pinnules apparently correspond MO.N .\1.VHI — 05 5 66 MESOZOIC FLORAS OF UNITED STATES. in position on the compound piiuia to the sterile ones shown in PI. IX, Figs. 3 and ('•>. PI. X, Fig. 5, gives a portion of one of these piiniules magni- fied two diameters to show the sori. PI. X, Fig. (1, represents, with Hke enlargement, a sterile pinnule with crenate margins showing the lateral nerves. PI. X, Fig. 7, gives, enlarged, two diameters, a fertile entii-(^ pinnide, to show the sorus. This fossil is not so abundant as Dickso7iia orcgonensis, l^ut it is on(» of the most common fcM'iis, ferns being mucli less connnon than cycads. It occurs most commonly, and in tlie best specimens, at locality No. 19, but is also foiind, not rarely and in good specimens, at locality Xo. IS. It occurs also at localities Xos. 1, 2, 7, and 16. Genus CLADOPHLEBIS Brongniart. Cladophlebis vaccexsis Ward n. sp. PI. X, Figs. S-12. LS96. CladopJilehis whitbiensis tenuis var. a Heer. Font.: Am. Joiirn. Sci., 4tli .ser., Vol. II, p. 274 (nomon.). 1900. PladopJtIehis ivhifhienfiis tenuis, var. a Heer? Font.: Twentieth Ann. Kep. r. S. Geol. Surv., 1S9S-99, Pt. II, p. .346." A fern of the Cladophlebis type of the general form illustrated in Cladophlebis ivhitbiensis, but not identical with the species of Lindley and Hutton, is found sparingly at some of the Oregon localities. It is always in a very fragmental condition, most of the specimens showing only l)its of ultimate pinnae with attac^hed pinnules. Xo fructification was seen. The constant absence of fructification on ferns of this type is a noteworthy feature. The largest specimens obtained give, at most, portions of ulti- mate pinnae so placed as to show that they were once attached to a common rachis. The size of the rachises and tlieir rigid aspect indicate that the plant must have attained considerable size and have been probably sub;ii'- borescent. There is a considerable variation in the pinnules in shajje and textin-e, which may perhaps l)e accounted foi- by different positions on the compound pinna, or different preservation. The forms graduate into one another in such a way that they can not wc^ll be separated. The piniui' aic long and slendei', with a rigid rachis. The pinnules have a "Professor KDntiiiiic docs not in his report incnticjii the iHiiincnci' «if lliis I'uirii in tlir Omvillc beds, and he rites a different li^nrc in Ilcci's work, but l>()tli figures refer to Heer's vnriely ii. In rorrespoiulcnce lie admits that the Oroville and Oregon forms are tiie same. — L. F. W. .Illv'ASSIC FL()1;A of I)()l(iFAS ('orNTV. ()UE(i. ()7 rather tliick coi-iaceou.s loxluro, atul, nol w ithstaiidiiig tli(> iVaiiinoiitary (■ondition of the specimens, ai-e often \r\-y well preserved. The)- are widest at base, attached li.\ the entire base, and often strongly falcate. In terminal parts th(>y ai-e sometimes shorter, l(>ss falcate, and more trianirulai- in form. They are in these parts more or less united, but usually are separate to the base. Th(>y are acute to subacute. The nerves are very distinct and are of the tyi)ical Cladophlebis type. The basal lateral neives are twice forked. Those hisiher up are less copiously i)ranclie(l, the highest being once foi'ked. Th(> bi'anches di\-erge strongl\- at first and aiv then parallel, so that they appear rather straggling. This plant agrees \n\ well with the form described by Ileer from the Jurassic of Sibeiia untler tlie nain(> Asjilcniinii wlutbicn.se Icnuc." It, is nearest his variety a. as figured on pi. xvi, fig. S. As these ferns are not specifically identical with the original I'cniplcri.s irhithicnsifi it seems best not to treat tlunn as a \-ariety of that species, and, as they are without fructification, it is safest to refer them to the noncom- mittal genus Cladophlel;)is. They are very much like the NeuroptcrU rcccntior (Phill.) L. & H., figured by Lindley and Hutton,'' but the pin- nules are usually much more acute than those given for the English plant. It may Ije, however, that the l)luntness of those of the latter is due to ■distortion or imperfect preservation, foi- a few of them are as acute as those of the Oregon fossil and have exactly the same shape. Mr. Seward has been kind enough to send me, along with the names of the plants figured, duplicates of the plates for his forthcoming work on the Yorkshire Jiu-assic Plants. The plates were not accompanied 1)>- descriptions, hence reference can be made only to the plates and figures. The work will be referred to as "Yorkshire Jurassic Fossils." ' In his paper entitled: "Notes on some Jiu'assic Plants in the Manchester Museum," which deals with some of these plants, he has, on pp. 8-11, brought together a large number of ferns of the ivldtbiensis type. In these " Fl. Foss. Arct., Vol. IV, Ft. II, pp. 38-40, pi. xvi, fig. 8. ''Fos. Fl. Gt. Brit., Vol. I, pp. 19.5-196, pi. Ixviii. •'Tlic work wus not received in Amerieii till after Professor Fontaine's report had been completed and the raanus<-ript and types sent by him to Washington in the spring of 1!)01. Copies arrived, however, heforo the report had been embodied in this paper, and I have made free use of it in working ont the synonymy and arranging the speeies in systematic order. A copy of it was placed in Professor Fontaine's hands, and there has been considerable correspondence between us relative to critical points. The final decision in all cases, as propo.sed or accepted by him, has been embodied in the terminology here introduced. In the i)rescnt; case he has decided to make a new species, and left the work of naming it to me. — L. F. W. 68 MESOZOIC FLORAS OF FXITED STATES. are includod the Pccoptcn's trhilbicnsis Brongn. and the Pecojitcris tenuis Schouw, descnl)ed l\v Brongiiiart. These he considers as identical with Pecopteris Williamson is. all l)eiiifi Todites. To these forms he gives the name Todites Williainsoni. In the paper on tlie plants of the Manchester Museum, pi. i, figs. 1, 2. he gives reproductions of two rather poorly pre- served specimens of Todites Williamsoni, that, with the exception of the smaller size of the pinnules, and the serration of their margins, agree pretty well with the Oregon plant. PL xxi, fig. 6, of his Yorkshire Jurassic plants, gives as this plant a fern that seems to have pinnules witli entire margins. This is exactly like the Oregon fossils. In the absence of fructification with the latter it can not be made a Todites notwithstanding this close similarity. PI. X, Fig. S shows parts oi several ultimate pinna* so placed as to indicate a former attachment to a common rachis. Fig. 9 gives a magni- fied pinnule to show the ner\'ation. Fig. 10 gives a portion of an ulti- mate pinna, with pinnides more remote than is common. Fig. 11 shows a pinnule of this enlarged, with nervation in detail. Fig. 12 shows the tip of an ultimate pinna, with shorter, more triangular, and more imited pinnules than is common in such situations. This is one of the rarer ferns and where it occurs l)ut few specimens were found. It occurs at the locality Xo. 19" most commonly, and is found also at locaUties Xos. 1, 2, 6, 7, 17, and 18. Cladopiilebis denticulata (Brongniart) Nathorst non Fontaine. PI. XI, Figs. 1-7. 1828. Pecopteris denticulata Brongn. [non Heer]: Prodrome, pp. 57, 198 (nomen). 1828. Pecopteris PhUlipsii Brongn.: loc. cit. (nomen). 1829. Pecopteris lif/ata Phill.: Geology of Yorkshire, pp. 148, ISO, pi. viii. fig. 14. 1833. Neuropteris tigata (Phill.) L. &. II.: Foss. Fl. Gt. Brit., Vol. I, j). I!i7, ])1. Ixix. 1834. Pecopteris insignis L. & H.: op. cit.. Vol. II, p. 69, pi. cvi. 1834. Pecopteris undans L. & H.: op. cit.. Vol. II, p. 103, pi. cxx. 1834. Pecopteris denticulata Brongn.: Hist. Veg. Foss., Vol. I, p. 301. pi. xcviii, figs. 1, lA, IB, 2, 2A. 1834. Pecopteris PMUipsii Brongn.: op. oil.. Vol. 1, p. 304, pi. cix, figs. 1, lA. 1836. PIdebopteris ? undans (L. & II.) Brongn.: op. cit.. Vol. I, p. 375, pi. cxxxiii, figs. 3, 3A, 3B a The specific name alludes to this locality (bed of Cow Creek). — L. F. W. .UKASSIC KI.OKA OF DOl'CI^AS ('OrXTY. OKK(i. 69 183G. Mitlioptins I'hiUijis'ii (Brongn.) (lopp.: Syst. Fil. Foss., p. ;>04. 1836. Al• 3(17. 1S3(). I'ohjpiidtUs undans (F. c^ ]F) CJdpp.: oj). cit.. p. 34."). 1848. Alilhopleris (It'iitlciildtii (Urongii.) Gr.pp.: Index Pal., XdnicncliUor, p. 23. 1849. Cladophleliis I'Kjutd (I'liill.) Brongn.; 'I'iililcaii, p. Id."). 186.'>. /'/m.s' I'xjata (PhiU.) liii.: Fanikniutcr dcr -Icl/.lwrlt . p. IFl. 186.5. PU'rift imignis (L. & IF) Ett.: op. cit., p. 114. 1874. CladophlrlHs hislgtvs (L. & H.) Schinip.: PaF V('g., Vol, II F ]). 50.5. 1876. Claih.phlrhis dnitn-ulata (Brongn.) XaMi. [non Font. J: Bidrag till Svcrigos Foss. Fl., Vii.xtcr fr. Riit. Form, vid Palsjo, ]). !'.>. 1N78. A.spltn'tuin. pffruxcliuirnsc Fleer: Fl. Foss. Arct., Vol. V, Pt. II (15eitr. /.. Fo.ss. Fl. Sibiriens), p. 3, pi. i, Hgs. 1, lb. 1882. Pteris frigida Heer: Fl. Foss. Arct., Vol. VI, Abth. II, Foss. Fl. Gronl.. Pt. I, pp. 3, 25, pi. ii, fig. 13; pi. vi, fig. 5b; pi. x, figs. 1-4; pi. xi, figs. 1-4, on, 6, 7a, 8a, 9-11; pi. xiii, figs. 2, 2b; pi. xvi, figs. 1, 2; pi. xviii, fig. lUb. 1882. Pterifi longipennis Heer: op. cit., p. 28, pi. x, figs. 5-13; pi. xiii, fig. 1. 1888. Cladophlehis denticulata Font, [non (Brongn.) Natli.]:' Potomac Flora, p. 71, pi. iv, figs. 2, 2a; pi. vii, figs. 7, 7a. 1896. Cladophlehis ste'rartiana Hartz: Medd. om Gronl., Vol. XIX, p. 231, ])i. xi, figs. 1, 2; 1)1. xii, figs. 2, 3. 1896. Asplenites ^^ sp. Hartz: op. cit., p. 231, pi. xi, figs. 3, 3a. Fragments of what was evidently a large fern of Cladophlebis type were foiuid sparingly at some of the localities. Considering the strength of some of the parts preserved, and their good state of preservation, the comminution of this fern is remarkable. Only sterile forms were found. The largest specimens show only small Ijits of ultimate pinnse. Some yield only scattered pinnules and fragments of pinnules. The pinnules seem to have had a leathery, firm texture. The rachis of the ultimate pinna? is strong and rigid. The pinnules are more or less falcate, some- times strongly so. The>' are attached by the whole of a somewhat expanded base. The larger normal pinnules are oblong linear in form, with lancet- shaped suijacute tips. Those in terminal parts are shorter, sometimes approaching a triangular form. The nervation is quite characteristic. The midnerve is strong and persists to near the tip of the pinnule, having '' Professor Fontaine dcsi-rihed tliis a.s a new species, but lie notes its reseniljlance to Pecopteris ilenlica- tata Heer non Brongniart from the Cretaceous of Greenland. Mr. .Seward places it in his synonymy of the Yorkshire plant wliich was described under that name by Brongniart in 18:j4, and which Nathorst seems to have been the first (ISTC), see synonymy) to refer to Cladophlebis. Mr. Seward, however, does not inchide Heer's plant in his .synonymy, and in mentioning it on p. 141 he seems to think that it was the same as Brongniart's, but lleer'.s was also called a new species. This makes an imforlunate confusion of names of closely related forms, which it is dilficull to make clear. — L. F. W. 70 MKSOZOIC i^LORAS OF UNITED STATES. less than usual of the Cladophelns charactei'. Tlic lateral nerves are strong and veiy distinct. They ai'e given off at an angle of about 45°, and fork neai' their insertion. The hranehes divei'ge strongly, then become parallel and turn outward to meet the margin under a large angle. One of the Ijranches may fork again, l:)ut they are mostly single. Some of the pinnules appear denticulate near their tips, but this may be due to lacera- tion and imperfect preservation. The larger pinnules are about 24 mm. long and (i mm. wide. This fine fern seems identical with the Pecopteris insignis of Lindley and Hutton." The larger pinnules are exactly like the form given bj^ Phillips for this species. '' Seward, in his paper on the Jurassic Plants in the Manchester IMuseum, identifies this and a number of other species with Cladophlebis denticulata (Brongn.) Nath., and on pi. iv he gives a figure of C. denticulata that agrees well with the Oregon plant. The pin- nules of this plant resemble so much those of the fine Danceopsis Storrsii, described farther on, that I am inclined to think that they are the sterile forms of that fossil. Lindley and Hutton described from the Yorkshire Oolite a fertile fern under the name Pecopteris undans,'^ giving a fructification like that of Dameopsis Storrsii. Seward, in the paper above quoted, p. 19, states that Xathorst had suggested that Pecopteris undans may be the fertile pinna of CUidophlehis denticidata, and further that an examination of sev- eral examples of Pecopteris undans enabled him to confirm Nathorst's opinion. The fructification, as given by Lindley and Hutton in the figure of f . undans, is strikingly like that of Danceopsis Storrsii and would make the plant a Dana^opsis. Seward gives no reason foi' connecting Clado- phlebis denticulata with this plant. There is no connection between Danceopsis Storrsii and the Oregon form referred to Cladopldchis denticu- lata, hence the latter nmst remain in the genus Cladophlebis. It is, however, significant to find that in both these cases the idea is suggested that the fructification of this Cladophlebis is that of Danseopsis. It should l)e stated that Seward does not in his remarks indicate the res(>m- blance of the fructification of Pecopteris undans to that of Danaeopsis. "Foss. Fl. Gt. Brit., Vol. II, p. 09, pi. cvi. '' PliiUips, (ii'()l()<;y i)f Yorkshire, .3tl ed., p. 2()(), li{;ii. 17 most complole specimen, whicli is a frag- .it of •inullimate pinna having several pinmiles. These are .lisL.rted and slickensid.Ml. The prcdnction of shckensides is not an unconnnon featnre in the Oregon phuits whenever they have a firm resisting textnre. The creep of th(. rock seems often to have taken pUice along tlie snrface of the plant, polishing it and obliterating the nerves. Fig. 2 shows an enlarged pinnule of this. Fig. 3 depicts a fragment <.f .an ultimate puma, havino- one entire pinnule that shows indications of denticulation toward ils tip'^ This pinnule is shown enlarged in Fig. 4. Fig. 5 gives a small fragment of an ultimate pinna with several pinnules of the more slender kind Fig. shows two fragments of ultimate pinn* with mutilated pinnules that occair toward the termination of the phma^. Fig. 7 gives a pinnule enlarged to show the nerves. The plant is most common, l)ut still rare, at locality No. 19, and is found also at localities Nos. 1, 2, 7, <). 12, U. 17, and 18. Clauoi-hlebis iiAiBrRNENSLS (Liudlev .^ Ilutton) Bron-niart ? Fl. XI, Fio;s. s-l(). ls:i(i. PecopWrls haihurnensis L. & II.: Fc.ss. Fl. Gt. Brit., V..1. Ill, p. 97, ,.1. clxxxviii. 1S49 Cladophhhls hmburnensis (L. & H.) Bron<,ni.: TahU-au, !>. 105. 1S6.5. Pterin haihurnensis (L. & H.) Ett.: Farnkniuter dor Jetztvvelt, p. 11-1. ISOq iJethoptens haihurnensis (L. & H.) Schiinp.: Pal. V.'g., \<)1. I, p. 56,5. isoo. ThluHJMla haihurnensis (L. & H.) Rac: Bull. Int. Aca.l. Sci. do (^acvie, Janvier, 1S90. p. ■Vl. A single specimen of a fern of marked Cladophlebis type was found at locality'^No. 1. It seems different from any of the other Cladophlebis of the Oregon Jurassic. The specimen shows only a fragment of an ulti- mate pinna carrying several pinnules. These agree so well with the tern called bv Lindlev and Hutton Pecopteris haihurnensis" that it is most probably the same species. The English fern is clearly a Cladophlebis and not" a Pecopteris. The amount of material, however, is not sufficient to justify a positive identification. It may possibly be an abnormal form of Cladophlebis vaccensis, but the entire aspect of the i)lant is different from that fern. The rachis is slender. The pinnules are slightly falcate and attached hx the entire, somewhat widened base. They are in form «Foss. Fl. Ot. Brit.. Vol. Ill, pp. (IT-itS, pi. clxxxviii. 72 :\rEsozoTC floras of united states. obloii placed in this genus. PI. XI, Fig. 11, gives one of th(» specimens, and l"ig. 12 an enlarge- ment of one of the pinnules. ("LADopin.Kius rEroPTf.T!on)Es Fontaine ii. sp. PI. XI, Fi<;.s. 13-15. An elegant fern, with finely-cut foliage, was found in two specimens at locality Xo. 1 . One of the specimens is quitc^ small and shows only a terminal portion of the peindtimate pinna that comes apparently from high up on the compound })inna. The other one, given in Fig. 29, shows considerable portions of the penultimate pinna> that seem to come from low down on the compound pinna. They are so placed as to show that they were once attached to the rachis of a pinna of superior order. By the small diminution of the pinnules they indicate that the ultimate pinuif, of which they formed parts, had a very considerable length. The rachises of these pinnse are strong, with their margins raised cord-like, and they have a cord-like rib running down their centers. These pinnse carry a number of pinnules that for a considerable distance maintain a very vmiform size. The pinnules are clo.sely ap[)roximate, sometimes overla})ping, and are falcate and 4-5 mm. wide. The pinnules are linear- oblong in form. They are about 1 cm. long, with obtuse or subacute tips. Their character gives the plant a marked elegance. They are set on the rachis at an angle of about 45°, and are more or less deeply incised into oblong or ovate, very obtuse lobes. The deepest incision goes about two-thirds of the way to the midnerve. These lobes or pinnules are strikingly like those of some of the Carboniferous pecopterids and the species is named from this resemblance. The nervation, however, is that of Cladophlel)is. The leaf substance is quite thick and obscures the nerves. Apparently the lobes, in lower parts of the compound pinna, pass into distinct pinnules, which may have crenate margins. In ter- minal parts, on the other hand, the lobed pinnules diminish to those with «F1. Foss. Arct., Vol. IV, Pt. II, (Beitr. z. Jura-Flora Ostsibiriens), p. 93, pi. xviii, fig. 4. 74 MESOZOIC FLORAS OF UNI TED STATES. crenate margins. In each lol)e there is a parent nerve that splits up into liranches. It may give off one or more pairs of lateral nerves that go off ver>- obliquely. The nerve gi'oup has a flabellate character. No fruc- tification was found. The plant is prohal)l>' a new species. It reseml)les veiy much the Pecoptens ohtuHifolin (Murray) Lindley & Hutton" and may be the Oregon representative of it. In the English fossil, however, the ultimate pinnae, corresponding to the pinnules of this plant, are much longer, and fructification was found. Phillips identifies the species of Lindley & Hutton with his Pecoptens c.rilis." Seward regards Pecopteris cxilis as a Klukia.'' PI. XI, Fig. 13. shows the specimen natural size, and Figs. 14 and 15 give enlargements to show details of nerves. Genus SCLEROPTERIS Saporta. SCLEROPTERIS OREGONENSIS Folltaiue 11. sp. PI. XII, Figs. 1-3. A single specimen of a fern was found of pretty good size and fairly we 1 preserved at locality No. 19. According to Saporta's diagnosis of the genus Scleropteris,'' this genus clearly includes the plant in cjues- tion. The specimen shows a considerable portion of a penultimate pinna, which bears fragments of a number of ultimate pinna', with pin- nules. The ultimate pinnse are very short, the longest of them not surpassing about 3 cm. The rachis is narrowly winged. The pinnules are rigid and have a very thick leaf substance that conceals the nerves, so that they generally are diflftcult to make out. The lowest pinnule on the upper side of the rachis is oblong, obtuse, and larger than the rest. It is crenate on the margins or has shallow obtuse teeth. The other, and normal pinnules, are not more than 5 mm. long and about 3 mm. wide. They are entire, oblong-elliptic in form, with very obtuse tips, and go off obliquely from the rachis, with a slight inclination toward the ends of the pinna^. They are decurrent, forming the wing. All the pinnules are constricted at base. The nerves are immersed appa- rent y in the leaf substance. They are composed of a parent nerve "Foss. Fl. Gt. Biit., Vol. Ill, pp. 15-16. pi. fhiii, fifr. 1. ''Geology of Yorkshire, 3d. od., p. 210, pi. viii, fig. l(i. 'Jura-ssie Plants in the Manchester Museum, p. 4; Jur. Fl. Yorkshire Con.st. p. 1,30. ''Pal. Franraise, Plantes Jurassiques, Vol. 1, pp. 364-36.5. JURASSIC FLORA OF DOIMILAS COlNrV. OHP:(i. 75 that goes off obliquely and which has several lateral nerves that depart under a small angle and seem to he single. The plant seems to l)e a new species. It has some resemljlance to Sderopteris Pomelii, hut the pinna' are shorter and the pinnules are shorter in {proportion to their width. They are also more closely placed and more ol)tuse. It is more like Heer's Dicksonia f/racilis," hut differs from that in the heteromor- phous l)asal pinnule and in the shorter ultimate j)inna'. It is near enough, however, to he regarded as perhaps a repi'esentative of the Sil)erian fossil in the Oregon Jurassic. PI. XII, Fig. 1, shows the only specimen found, Fig. 2 a pinna enlarged, and I'ig. 3 an enlarged normal pinnule. Genus RUFFORDIA Seward. RuFFORDiA GfEPPEKTi (Duiiker) Seward.'' PI. XII. Ficrs. 4-S. 1S44. CheUantliitcs Gappertl Dunk.: Norddeutscli. \Valderthon, Prograniin der hoheren Gewerbscliule in Cassel, 1843-1844, p. 6. 1846. Sphenoptei-is Gccppeiii Dunk.: Monoi);r. d. Norddeutscli. Wcaldenhildun,^, p. 4, pi. i, fig. 6; pi. ix, figs. 1-.3. 1S4G. Sphenoptrris Ilartleheni Dunk.: op. cit., p. 4, pi. ix, iig. 9. 1846. Sphenopfo'i.s long'ijoJiit Dunk. [:non PhiU. nee Feistni.]: o]). eit., p. 4, |)l. viii, fig. 6. 1851. tiphenopteris adianiifrons Ett. : -Jahrli. d. k. k. Geol. Reich.sanst., Jalu'g. II, p. 1.57. 1852. Sphenopteris JugJeri Ett.: Beitr. z. Fl. d. Weald enperiode, p. 15, pi. iv, fig. 5. 1870. Sphenopteris Auerhachl Trautsch.: Der Klin'sche .Sandstein, Nouv. Mem. Moscou, Vol. XIII, p. 207 [19], pi. xviii, fig. 5. ISSl. Sphenopteris valdensis Heer [in part]: Fl. Fo.s.s. du Portugal, p. 14, ]>!. xv, fig. 11. 1889. Sphenopteris sp. Yok.: Joiirn. Coll. Sci. Ini]). I'niv. Japan, \o\. Ill, Pt. I, p. 34, pi. xiv, figs. 13, 13a. 1894. Ruffordia Ocepperti (Dunk.) Sew.: Wealden Flora, Pt. I, i)p. 7C), 77, pi. iv; pi. v; pi. X, figs. 1, 2. " Fl. Foss. Arct., Vol. IV, Pt. II, p. 92, pi. xvii, fig. 3. I' Mi-. Seward (Wealden Flora, Pt. I, p. 76) includes in his synonymy of this species the Sphenopkris Phil- lipsii of Mantell, published in 1833, or eleven years earlier than Dunker's Cheilanthites Gapperti, but still pre- fers Dunker's specific name on the ground that Mantell's specimen was a fragment "so .small that its real nature must remain uncertain," but without saying whether he had seen it. As the use of Mantell's name in the synonymy would change the combination, oui- only course is to omit it, as Fncniiies mruntiis L, it H. was omitted from the synonymy of M archantites erecliis (Bean) Seward. — L. F. W. 7(3 MESOZOIC FLOKAS OF UMTKI) STATKS. Phillip.s" fiives a fifiviic, witli no ck'sniplion, of a singular plant that he calls Splicnopteris Jncjleri!' This, in f,^cncral aspect, agrees so well with one found in Iwo specimens, one each at localities Xos. 12 and 7. that I think they belong to the same species. The Oregon plant has apparently no foliage, hut is composed of a thread-like stem, which was prol)ahlv succulent, as no vascular tissue shows in it. This rachis branches in an irregular straggling manner, widely diverging after each branching. It bears, irregularly placed, short branches, or contracted pinnules, that sometimes are forked and sometimes single. Their summits are expanded into elliptical forms, so that the branch is club- like in shape. Xo nerves are visible. Each expanded tip contains a rounded Iwdy, depressed in the center, that looks much like a sorus. The preservation is not sufhcient to show its true nature, but, if it is a sorus, it probably had an indusium. This plant much resembles Schenk's Acrocarpus cuneahis,' from the Rhetic, but the segments are much smaller than those of that plant. PI. XII, Figs. 4 and 7 give different fronds; the latter is a small fragment, but more distinct. Fig. 5 gives an enlargement of a portion of Fig. 4, and Fig. 6 a pinnule still more enlarged. Fig. 8 shows a por- tion of Fig. 7 much enlarged. Genus ADIANTITES Goppert. Adiamites Nympharum Heer? PI. XII, Figs. 9-11. 1876. Adiant'des Nympharum Ileer: Fl. Foss. Arct., Vol. IV, Pt. II (Jura-Fl. Ostsibiriens) , p. 93, pi. xvii, fig.s. 5, 5b. Very imperfect specimens were obtained, one each from localities Nos. 2, 7, and 19, of a plant that resembles Heer's Adiantites Nynipha- rum. The pinnules are always too much mutilated to show their true form and onlv small bits of ultimate pinnse were obtained. All that " Geology of Yorkshire, 3d ed., p. 218, lign. 40. '' Pmfpssiir Fontaine, bpforo receiving the text of Mr. Seward's .lurnssic Flora of tlie Yorl^sliire Coast, wa-s inclined to identify tlie Oregon plant with Hi>henoplfnK Jugkri Ett. on the strength of its resemblance to Phillips'.s ligure, ImtJn view of tlie fact that Mr. Seward says (p. 1.3:? ) that this ligiire "does not do justice to the original," and also because .Mr. Seward refers all the plants called SiAniopleris J wjleri Ett. to Rufonlia Gapperti, it was decided that our plant must belong to that species. — L. F. \V . <■ Foss. Flor. der Oren/.schichten. pp. 134, 13.5, pi. xx, figs. 9-12. JlKASSir FLOKA OF lH)l(iLAS CorNTY, (niKd. 77 can be said is tlint \]io piaiit may be Heer's fossil. The piiuniles are narrowed wediic-sliapc to llic base and seem to have dentate niariit,'niart. Forms like Ta-niopteris are more common in tlie Oregon Jurassic flora than the ferns with smaller pinnules. 1 shall use the distinction suggested by Xathorst as an essential one between Tamiopteris and the vmsegmented Xilsonias that, in shape, so much resemble Ta^niopteris. This distinction is that the lamina of Taniiopteris is attached to the side of the midrib and in Xilsonia to the upper surface. This feature causes a Tjeniopteris to show a distinct midrib, whether the upper or the under surfac(> be presented uppermost. In the case of Nilsonia, however, when the upper surface is seen uppermost there is no visible midrib or axis. The nerves l^elonging to the lamina on opposite sides of the axis meet in a raised cord in the center of the position that would be occupied by the midrib if it were shown. But if the lower surface of a Nilsonia be presented uppermost the axis or midrib is seen, and, therefore, while the absence of a distinct midrib may be taken as showdng that the plant is a Xilsonia, yet, in cases where a midrib is shown, one can not be sure that the plant is not a Nilsonia with its imdei' surface presented uppermost. There are in the Oregon collection a number of leaves that in their shape are like TaMiiopteris. The}- are, however, never seen with their lamina" di\-i(led or segmented. They show no mitlrib, but have their lateral nerves meeting in a raised cord that occupies the central line of the position that would l)e occupied by the midrib if it were present. These leaves have uniformly in their laminae a thin texture. In some the lateral nerves are always single and show no thickening toward their bases near their insertion on the central cord. Others, with the same unchanged thickness in the lateral nerves, have them rarely forked, but in such way as to show that the essential character here, too, is an 78 MESOZOIC FLORAS OF UNITED STATES. iinforked condition. .Xotwillislanding the entii'c nature of tlie laminae of these leaves I shall class them as Xilsonia. There are other ta>niopteroid forms in this floia tliat always show strong prominent midril)S. They have a ver>- thick leaf-substance that is often slickensided from the creep of the rock alono- their surface. The lateral nerves are, as a rule, forked, and are much stronger, or thickened, near their insertions. These nerves go off approximately or quite at right angles, and are not curved, while in the Xilsonias they go off obliquely and curve in a characteristic wa>- to meet the margin. There is one exception to this character found in the form Tmiiopteris orovillensis, to be noted further on. This has the niidiib of Tjeniopteris and the ateral nervation like that of the Nilsonias. The question might arise, Is not this really a Xilsonia with its under surface presented uppermost? The leaves that I group as Ta^niopteris very often occur .with a number together in the same hand specimen of rock, as if they grew in clusters. Those regarded as Xilsonia are usually found in isolated fragments. T^NIOPTERIS OROVILLENSIS FoiltaiuC. \ PI. XII, Figs. 12-17. 1896. Txniopieris ororillensis Font.: Am. Jouni. Sci.. 4th sor.. Vol. II, p. 274 (nomen) . 1900. Tseniopteris orotnllensis Font.: Twentieth Ann. Rep. V. S. Geol. Surv., Pt. II, 1898-99, p. 348, ])1. hi, figs. 2-4. This Ta^niopteris, so al:)undant in the Oroville Jurassic flora, is found also at some of the Oregon localities. It is not so al^undant and widely diffused as some of the other species of Tfeniopteris. The nerves are fine but very distinct. They have the character, as men- tioned above for Xilsonia, of being always single, not varying in strength from their insertions to their ends, and of going off at less than a right angle, and of curving in their course to meet the margin. The leaves, however, have a much thicker and more coriaceous texture than those regarded as Xilsonias, and the midrib is strong. The size of these leaves varies a good deal. The most common forms have the sizes seen in the Oroville specimens, but occasionally fragments are found indicating a somewhat wider leaf than any seen in the Oroville plants. .HHASSIC FLOliA OF lM)r(iLAS COIWTV. OKKC. 79 In tho (lose rip lion of this spoeies cjivrii in llu^ Twentieth Annual lleport of tlie United States Ge()loo;ifal Survey, Pt. II (p. ;US), tho stat(Mnent, by mis])rint, is made that "the fronds vai'\- in length from 1-4 cm." Foi' "length" read width. The fragmenls .seen indicate that .some of the Oregon leaves may ha\e reached the widtli of o cm. PI. XII. Fig. 12, shows the upp(>r part of one of the larger leaA'os, and Fig. I'A a portion of this enlarged. Fig. 14 represents the basal portion of ont^ of the nai'rower leaves that was evidently (juite long. It is distorted by pi'essiu'e. Fig. lo gives the middle poi-tion of on(> of the leaves of nuMlinm size. This, too, is distorted. Fig. Ki sliows a leaf nearly entire, and Fig. 17 a portion of this enlarged. The plant is most common at locality No. 2, wdiere it is (juite abun- dant. It is also found at localities Xos. 1, 7, 14, and 19. T^MOPTEKLS MAJOR Filldlov illld Huttoii. PI. XIIl. Ki.^.s. l-;5. 1833. Tctnlopteris major L. & H.: Foss. Fl. Gt. Brit., Vol. II, p. 31, pi. xcii. 1836. Asjndites WJUiamsonifi Gopp.: Syst. Fil. Foss., p. 353. 1843. Fterozamites major (L. & II.) Fr. Br. in Miinstor: Boitr. z. Petrefacten-Kunde, Vol. II, Heft VI, p. 29. 1865. Txniopteris WiUiatrisonis (Gopp.) Zign. : Osserv. sullo Felci Foss. dell' Oolite, p. 39. 1869. Macrotxniopterifi major (L. & II.) Schiinp.: Pal. V('<;;., Vol. I, p. 610. Lindley and Ilutton have described wdth the name Tceniopteris major " a fern that agrees so well with one foimrl in the Oregon Jm-assic that there can be no doubt that it is the same. The dimensions of the leaf and the size of the midrib in the two are the same, but there is some difference in the latei'al nerves. In the figure given l)y Lindley and Hutton of the English plant these nerves are represented as verv slender, and not varying in thickness from their insertions to their ends. They are also given as forking very copiously, especially at the margin of the leaves. In the Oregon fossil the lateral nerves are decidedly stronger near their bases, where they are attached to the midrib, than they are near the margin of the leaf. Owing to the great thickness of the leaf substance, and the fact that the plant is generally slickensided, it is difficult to see them distinctly. This is especially true of the parts near the margin of the leaves, for they have suffered « Foss. Fl. Ot. Brit., Vol. II, pp. 31-32, pi. xcii. 80 MESOZOIC" FLORAS OK rMTED STAPHS. inoi-o here llian elsewhere. They, liowever, do not seem to show such I'opious forking here as the fossil of Lindley and Hiitton. The follow- ing seems to l^e their character: The lateral nerves go off at a right angle, fork in a dichotomous maimer, mostly at two-thirds of the dis- tance from the midrib to the mai'gin. Near the margin one oi- l)oth of the branches may fork again. The forking occasionally occurs nearer the insertion of the nerves. The branches, after abruptly diverging, become more or less parallel. The lateral nerves of l)oth the English and the Oregon plants form a very characteristic featiu'e, and in the Oregon fossil readily distinguish the plant if they are visible. There is in the size and shape of the leaves no obvious general distinction between this plant and T. 42, pi. xxxvii, lifr. 2." 1 (ind no such designation in Sternberg's work, and the language used by Mr. Seward seems to be borrowed from Brongniart. In his Prodrome, p. 02, he .says: '■Titniopleris vittata: Scitaniineanim jolhnn t Sternb., fasc. 3, pag. -12. tab. :^7, fig. 2: Filicites ? ejusd. fasc. 4 (in indict- iconum)." Practically the same entry is .irKASSIC FLORA OF DOFCiLAS COFXTY. OREG. 81 viii, iiir. •>. 1829. Scolopoidriinii s-i>li/(irlinii Pliill.: (Jcoloj^y ol' Yorkshin', |). 147, pi. 1831. TifniiopUris ritfdiii liion.u'ii.: Hist. Voj;. Foss.. Vol. I, ]>. -'(i.;, pi. l.xx.xii. figs. 1, 1.1, ■-• 4. 183(). Aspidit(S Ticniopttris (iripj).: Sy.-;ii.) Schiinp.: Pal. Vi'^'-. Vol. 1, p. liOT. Numerous specimens of a narrow Tieniopteris were obtaiiHMl at some of the Oregon Jurassic localities. They ag;ree very closely with 7'. riltdtd. and there is no doubt that they l)elong to this species. The specimens in shape and size resemble a fjood deal the nari'owcr forms of T. own'llcnsis, and when the lateral nerves are not visil)le can not well be distinguished from that fossil. Unfortunately, in this case also, as in that of T. nuijor. the leaf substance is so dense and the speci- mens are so much polished b>' slickensides that it is generally difficult to see them distinctly. This fossil is generally narrowly elliptical to linear ribbon-shaped, narrowing gradually to the base and apex. Tlie midril) is propor- tionally very strong, and is prolongetl into a long stipe, indicating that the frond was simple. The lateral nerves are slender and rather remotely placed. They go off at nearly or quite a right angle, and go parallel to one another to the margin. They appear to be mostly simple, but are sometimes forked. The forking, however, takes place in no regu- lai' way and in no partictdar position, Init seems, as it were, accidental. There is a considerable variation in the width and length of the leaves. With some doubt I unite with this species the form depicted in PI. XIII, Fig. 0, found in only a single specimen. This differs from the made in the s^-nonyiny of this spccie.s in his Hist. Veg. Foss., Vol. I, p. 2fi3. Tlu! reference is probably to the Frencli edition, a.s there are only 40 pai^es in fascicle 3 of the original (ierniun edition, 1823. On page 37 of that fascicle Sternberg mentions tlie plant figured in pi. x.x.xvii, fig. 2, and says: "Fig. 2 scheint eher ein Blattstiick einer .Scitaniinea uls ein Farrenkraut zu seyn." He also states here that this specimen came from Stone.sfield. On page 39 of the .same fascicle he enters the plant systematically under the general head ''Fili- cites" as " Phjllites scUamineujorinis," referring to the same plate and figure. This name also occurs in the index iconum. It does not occur elsewhere in the work, but is the only binomial apix-llation that lie applied to the plant. If it were certain that this specimen from Stonosfield belonged to the same species as the York- shire forms that Brongniart called Tmnio-pteris vitiata, the proper name for the species would be that of Stcrn- berf , wliich antedates Brongniart 's name by five years. An examination of Sternberg's colored figure, liowever, makes tliis doubtful. The character of tlie neri'ation is ob.scured by the effort to be artistic, and not eniaigh (if the leaf is shown to be certain as to its shape. Certainly nothing .short of a comparison of the type specimen could positively decide tlie <|uestion. This does not seem to have been done, and I therefore omit all reference to it from the synonymy of Txniopteris vittata. — L. F. W. MON XLVIII — 0.5 6 82 MESOZ(n(^ FLOl^AS OF UNITED STATES. ordinal-}- form of Tdniopteris vittain in its unchanging width, through- out the specimen, indicating a very long and nai'row leaf which must have been ribbon-shaped. It shows no lateral nerves. PL XIII, Fig. 4, gives the lateral part of a leaf of the larger size, which shows a considerable part of the stipe. Fig. 5 shows the basal portion of a leaf of medium size. Fig. 6 represents the abnormally long antl narrow leaf above referred to. Fig. 7 is a view of a fragment that shows the nerves, and Fig. 8 is a portion of this enlarged. This plant is most common at locality Xo. 7, where it is abundant and occurs wdth Tirniopteris mnjor, several of each sometimes being found in the same hand specimen of rock. It is found also at localities Xos. 1, 2, 4, 16, 17, and 19. T^XIOPTEHIS ? OREGONENSIS FoiltaillO 11. sp. Pi. XIII, Figs. 9, 10. At locality No. S was founti a single imprint, with its reverse, of a plant of tseniopterid character. It is 5 cm. long and 1 cm. wide, with no marked change in width, indicating a leaf of considerable length in proportion to its width. The midnerve is proportionally strong and no lateral nerves are shown distinctly, as the leaf substance is rather thick. Traces of them, however, can be made out. They seem to go off at right angles to the midrib. It could not be determined whether they branch or not. Possibly this may be the same species with the narrow form placed in Tceniopteris vittata, but the leaf is much smaller. This fossil is much like the plant given by Heer as Cycadites sibiricus," which is evidently not a Cycadites, but rather some form of Taeniopteris. As the amount of material does not suffice to fix the character of the plant, its position must be left doubtful. PL XIII, Fig. 9, represents the only specimen seen, and Fig. 10 a part of this enlarged. Genus MACROT.ENIOPTERIS Schimper. Macrot.«;niopteris californica Fontaine. PI. XIV, Figs. 1-4. 1896. Macrotxniofteris californica Font.: Am. Journ. Sci., 4th ser., VoL II, p. 274 (nomen). 1900. Miicrot^ n'loptcvlti californica Font.: Twentietli Ann. Rep. U. S. Geol. Surv., 1N9S-99, Pt. II, p. 349, pi. liii, fig. 1; pi. liv., figs. 1, 2. "Fl. Foss. Arrt., Vol. V. ?l . II. pp. 10-17. pi. iv. fij. 1. JURASSIC FLORA OF DOUCxLAS COUNTY, ORFG. 83 Three specimens of a fine taeniopteroid plant were found at locality No. 7. It agrees so well witli Macrotcenioptcris catif arnica of ihe OioNJlle flora tluil I have no hesitation in regarding it as the same. The Oregon specimens are l^etter preserved than those found at Oroville and show the margins. The plant must have had a very thick leather-like texture, as it leaves on the rock a decided film of coal. The maximimi width shown is 6 cm. The greatest length seen is 14 cm. and this in a mere fragment. The midnerve is strong. The latei-al nerves are seen with difficuUy. They ai'e mostly imforked and parallel to one another. They are rather remote, being about three-fourths of a millimeter apart. Some are forked at rather more than half way between the midrib and margin. The fork- ing, however, appears, as it were, accidental and not essential. The plant resembles Saporta's Tmiiopteris superhn," but the nerves are more remote and not forked at the l)ase, as in Saporta's fossil. PI. XIV, Fig. 1, gives the basal portion of a leaf. Fig. 2 represents a considerable part of a leaf from some distance aljove the liase, indicating a leaf of great length. Fig. 3 gives a fi'agment in which the nerves are visible, and Fig. 4, a portion of this enlarged. Family MARSILEACE.E. Genus SAGENOPTERIS Presl.* Sagenopteris Gosppertiana Zigno. PI. XIV, Fig.s. .5-11. 186.5. Sagenopteri.1 Gceppertiana Zio;ii.: Osserv. sulle Felci Fos.sili delJ'Oolite, p. 30. 1865. Sagenopteris Brongniartlana Zign.: Loc. cit. 1865. Sagenopteris Brauniana Zign.: Loc. cit. 1865. Sagenopteris rofiiridata Zign.: Loc. cit., p. 35. 1868. ? Sagenopteris Gappertiana Zign.: Fl. Foss. Form. Oolith., Vol. I, p. INN, i)i. xxi, figs, la, It), 2-.5; ])!. xxii, figs. 1, 1. 1900. Saqenopteris Nilsoniana (Brongn.) Wani: Twentieth Aim. Rep. U. S. Geol. Siirv., 1898-99, Pt. II, p. 352, pi. Ivi, fig. 1; pi. Ixvii, fig. 2. «Plaiitos Jura.ssi(iucs, Veil. I, pp. 4:W-441, pi. Ixi: pi. Ixii, lip;. 1. ''Potoiiic, who is working \ip the Pteridopliytii for Kii{;lfr and Prantl'.^ Natiirliclic I'llanzi^nraniilicMi, classes this genus in the Marsileacea', following the views of Heer, Zigno, Fei.stinantel, Nathoist, and otliers; but Count Solms-Laubach sees reason to doubt the correct ne.ss of this, and Seward treats Sagenopteris as a fcm, classing it now (Jur. Fl. Yorkshire Coast, 19(X), p. Ifil ) in the family Polyi)odiai-eie. alllumgli in liis Wealdcn Flora, I,S94, p. 120, he placed it in the SchizeaceiP. — L. F. W. 84 MESOZOIC FLORAS OF l^NITKD STATKS. Zigno has described from tlie Lower Oolite of Italy a fossil with the name Sagenopteris Gappcrtinua" that exactly resembles a plant occurring rather abundantly at some of the Oregon Jurassic localities. He gives a number of figures which show that the plant varies a good deal. Most of these variations can be seen in the Oregon specimens. The plant has a well-mai'ked character. The lai-gest Oregon leaves have a length of 9 cm. and a width in the widest portion of 35 mm. The leaves vaiy nuich in size and in other points. They are all decidedly ineciuilateral and tend mostly to assume a spatulate shape, widening toward iheii- cuds. Occa- sionally a leaf shows a narrowing at the tip, so that it is subacute. These seem to be the central leaves of a group. But most of them are very oljtuse at their ends and rovmded. These are rounded off toward their bases elliptically. They are apparently the lateral leaves of a group. Some of the obtuse leaves are narrowed gradually to their base, giving the base a prolonged wedge form. None were seen attached. The mid- nerve show\s consideral)le variation. In the leaves with prolonged wedge- shaped bases it is carried two-thirds of the length of the leaf. In those with the most marked inequilateral forms and elliptic bases it is not so prolonged, going, at most, one-third of the length of the leaf; in some it is hardly at all developed. The secondary nerves are not dis- tinct. They are very closely placed and slendei', anastomosing so as to form long meshes. The branches in anastomosing meet at very acute angles. One form that seems to belong to this species is abnormal in l)eing short, broad, and l)roadly elliptical, with a rounded base and hardly any development of midrib. Another is abnormal in being very small. It is only 5 cm. long. This is l)ut slightly inequilateral and may l)e a form of Sagenopteris paucifolia. This is proportionally not smaller than the fossil given by Zigno, pi. xxi, fig. 2, but it is narrower and proportionally longer than Zigno's plant . From an inspection of the more abundant and better material obtained at the Oregon localities, I am convinced that this plant is the one fomid in the Oi'oville flora and regarded as Sagenopteris Nilsoniiina'' {S. rhoifolia Presl.). PI. XIV, Fig. 5, represents a normal leaf that is strongly inequilatei-al, with a base that is rounded off m an elliptical forai. I'ig. 6 gi\'es the ter- "Flor. Foss. Form. Oolitli., Vol. I, pp. 188-190, pi. .\xi, figs. 1-5; pi. xxii, figs. 1, 2. ^Twentieth .\mi. Hep. U. S. Geol. Surv., Pt. II, 1900, p. 352, pi. Ivi, fig. 1 ; pi. Ivii, fig. 2. JURASSIC FLORA OF 1)01'(;LAS (OrXTY. ORFO. 85 iniiial poi-tioii of what was pi-ohalily a cpiiti'al loaf. This is less iii('((uilat- ci-al than usual and is su1ku'iiI(> at the tij). Murs at localities Nos. 1, 7, and Ifi. Sa(jkx()1'tekis paucif-olia (Pliill.) Ward n. coinl)." PI. XV, Figs. 1-8. ISL'U. Picopti r'm jxnic'ijolld Pliill.: Geology of Ydrksliire, [>. 14S, ])1. viii. fig. S. IS'Ji). Pfcoptcris longijolld Phili. (iion Broiigii.): Op. cit., p. ISl), |)i. \\\\. fig. 8 (probably nii.sprint for P. pduci folia). is:!(). Glossopferls PhilUpsii Broiign.: Hist. Vi'g. Foss., Vol. I, j). 22."), |)1. I.xi l)is, fig. .5; |>l. Ixiii, fig. 2. 18:^5. Otopteris cunniUi L. & H.: Foss. Fl. Gt. Brit., Vol. II, p. 20.3, pi. civ. 18.36. Acrostichifes Phillipnii (Brongn.') Gopp.: Syst. Fil. Foss., p. 286. ^'This is the plant tliat Mr. Seward (Jur. Fl. Yorksh. Coast, p. 162) calls "Sagenopieris PhilHpsi (Brong- iiiait)." He heads tlie synonymy with the reference to Brongniart's Hist. Veg. Foss., p. 225, where he describes Gidssopleris PhiUipsii, but dates it 1828. Tliis, it is tnip, is the date of tlie first volume, but it is well known that the work wsis published in parts, and it is very difricult now to ascertain the dates of the fa.scicles on account of the vicious h'?il)it of destroying the covers in binding sucli volumes. M. Kenr Zeiller has been to great pains to determine the dates of the parts of this work and lias been sufficiently successful for all prac- tical purposes. His results may lie found in the text to his monumental work on the flora of the coal basin of Valenciennes. (Ministere des Travaux Publics. Etudes des Oites Mineraux do la France. Bassin Houiller de Valenciennes. Description de la Flore Fossile, par R. Zeiller. Texte. Paris, 1888. Index bibliograph- ii|ue, pp. TOUT, cf. p. 703). From this it appears that page 22.5 was in the 5th livraison, issued in 18.30 together with pi. Ixi bis, containing the first figure. PI. Ixiii, containing the other figure, was included in the Oth livraison, wliich appeared in 1831 or 1832. Now as Brongniart puts Pecojilrris jmucijolta I'hill. in his synon3-my, it might have been seen tliat Phillips's work had then appeared, and that the date must have lieen later than 1829. This,also shows that Brongniart regarded Pliillips's plant as the same as his, coming as they all did from tlie same locality. He, of course, had no right to change the specific name, but the rules of nomenclature were very loo.se in those days and .still are with .some authors. There seems to bi' no escape from recognizing Phillips's name. — L. F. W. ''Although (ioppert expressly excludes Brongniart's forms from his synonymy, and deals only with (hose of Lindley and Iluttcm, he retains Brongniart's specific name, which is incomplete without his authority attached. Goppert's reasons for separating the forms are no longer considered valid. Mr. Seward includes the jUpiditts Xilsonianus, to which Gi'ippcrt r(>fers Brongniart's forms, in his .synonymy, but (hey form only a part of it, and that name relates to Brongniai'l's Fdicites NiJ>!(mianus, which Prcsl later renamed Sii<]enf>/>- teris rhoilotia (cf. Twentieth Ann. Kep. U. S. Geol. Surv., Pt. II, lOOO, p. 3.52). 1 therefore omit it from (he synonymy of this species. — L. F. W. 86 MESOZOIC FLORAS OF UNITED STATES. 1S36. Adiantites irregularis Gopp.: Op. cit., p. HS5. 1838. Cydopteris cuneata (L. & H.) Presl [non (Gopp.) Ung. noc Carr.]" in Stern- berg: Flora der Vorwelt, Vol. II, p. 135. 1838. Tifnloptcris PhiUipsii (Brongn.) Pre.sl in Sternberg: Op. cit., ]). 140. 1843. Sagenopteris ? cuneata (L. & H.) Morr. : Cat. Brit. Foss., ]). 20. 1849. PJiyUopteriti PhiJUpsii Brongn.: Tableau, pp. 22, 10.5. 1865. Gymnogramme cuneata (L. & H.) Ett.: Farnkriiuter der Jetztwelt, p. 70. 1865. Gymnogramme PhiUipsii (Brongn.) Ett.: Op. cit., p. 71. Forms that agree exactly with Sageno])icris jxtucifolia are not rare at some of the Oregon Jurassic locahties. The leaves vary a good deal in size but are pretty constant in general shape. They are narrowly ellip- tical, sometimes so narrow in proportion to their length as to be linear- elliptical. They narrow gradually to their liases and tips, so as to be lancet-shaped at the end and wedge-shaped at the base. The narrowing in some bases is more gradual than in others, so as to give the base a pro- longed wedge shape. The nerves are not so closely or so copiously anas- tomosed as in S. Goeppertiana and are not so fine. The leaves are decid- edly less inequilateral than in that species. Phillips '' has pointed out that the nerves depicted in the figure of Lindley and Hutton ■■ are not correctly given. Certainly no such nerves occur in the Oregon plants. The mid- nerve is more distinctly defined than in S. Gceppertiana, and is prolonged farther in the leaf. Seward "^ described two forms of this plant. One, the form called originally Otopteris cuneata, is not found in the Oregon col- lections. The other, given in fig. 8, agrees well with some of the Oregon plants. PI. XV, Fig. 1, represents one of the largest leaves. It is nearly entire. Fig. 2 gives a leaf with a base more elliptic in form. Fig. 3 shows the basal portion of one of the smallest leaves seen. This plant is nowhere abundant, l)ut is pretty widely difi'used. It is most common at locality No. 2, but occurs also at localities Nos. 1, 7, 14, 18, and 19. "Unger (Synops. PL Foss., p. .56) referred Gc'>ppert's Carboniferous species Adiantites cuneatus (Syst, Fil. Foss., p. 226) to Cyclopteris, and Cari-uthers (Quart. Journ. Geol. Soc. London, Vol XXVIII, August. 1872, p. 3.54) named a new species Cyclopteris cuneata from the Carboniferous of Queensland. — L. F. W. ''Geology of Yorkshire, 3d ed., p. 203. cFo.ss. Fl. Gt. Brit., Vol. I, pi. Ixiii. ''Notes on .some .Jurassic Plants in the Manchester Museum, Manchester Memoir's, Vol. XLIV, Pt. Ill 1900, pp. 11-14, pi. iii, figs. 7, 8. JURASSIC FLORA OF DOFO.FAS COrXTY. OUF/l. 87 SaGENOI'TKHIS (iKANDIKOI.lA F.mtlUllc n. SJ). r\. XV. Figs. 4, 5. A singlo imprint and its roverso of what nuist have Ixmmi a very larjic leaf was ol)taine(l from locality No. 7. It is evidently a Sajienopteris (luite different from either of the two above descril.(>d. rnfortimately the fragment is too small to show the size and shape of the leaf. It is a portion of the npper pai't of a leaf and shows the mar or less falcate. The tips are obtuse to subacute. They maintain the saiue width fi'om the 88 :vn-:soz()i(' floras of fnitfi) states. l)ase to some distance above it and ace attacticd hy the entiiT base. The midnerve of tlio pinnules is distinct and maintained to near their ends. The lateral nerves are not shown, but apparently in their place occur two closely approximate parallel, paired bands, that, to the unaided eye, look like \'ery thick nerves. The lens shows that these bands are concrete rows of minute punctifoi'm bodies that seem to be the capsules. These paii-ed bands or rows leave the midnerve at a large angle, and, in tlie case of two adjacent bands, go from a common point. They go nearly parallel to one another, so as to meet the margin nearly at a right angle. The paired bands seem to follow lateral nerves, one member of the pair being on each side of a nerve. This form is much smaller than the two Triassic fossils. Dnnaopsis marantacea Heer, and D. Rumpfii Schimp. As stated l)efore, tlie genei'al aspect of this plant is strikingly like that of Cladophlehis dcmticulata. PI. XV, Fig. 6, shows the most complete specimen. Fig. 7 shows an enlarged pinnule, and Fig. 8 a portion of this still further enlarged. Fig. 9 gives another enlarged pinnule to indicate the distribution of the capsules. I name this plant for Mr. Storrs as a slight recognition of his \-aluable work in collecting these plants. Order EQUISETALES. Family EC^UISETACE.E. Genus EQtTISETUM Lininpus. Eqiisktum ? sji. Foiitiiiiip. PI. XV, FUr. 10. The collections show no remains of alga? and no unequivocal Equi- setum. There are a few fragments of small stems that may be some form of Equisetum, but they are decorticated and of doubtful position. These fragments are about 5 mm. in width and sometimes 5-7 cm. long. They show ribs or cord-like cylindrical raised lines, running longitudinally, that do not seem to be any form of nerve. No articulations are indicated and no sheaths or teeth were seen. At most these forms are merely suggestive of Equisetum. They are among the I'arest of fossils. Such a form occurs at localities Xos. 2, 7, and 19. .URASSIC FLORA OF DOFCiLAS C'OFNTY, OREG. 89 Class GYMXOSl^KKM.K. Order CYCADALES. Family CYCADACE.E. Cicmis PTILOZAMrrES Natlior.st. Ptilozamites Leckexbyi (Bean) Nathorst. PI, XVF FiKs. 1, ■-'. Isr.S. Cfen;.. Lrckruhji Hoan in L.'.^k.nhy: Quart, -lour. (ionl. Soc. London, V.>1. XX, p. 7S, pi, X. fi^s. la, 11). \mAU)donfopt,n. .' Lcckenhip (Bean) Zign.: Fl. Foss, Form, Oolitl,,, Vol, L ].. 11 1. ISC') Cycadopteris Lecl-e.nhyi (Bean) Schimp.: Pal. Vi'g., Vol, 1. p. 4s,, ISSO. PtihzamUes Leclcenhyi (Bean) Natl,.: (MV. K. Sven.sk. Vet.-Aka.l. l-r,rh.,Vol. XXXVn, No. 5, pp. 65, S3. Phillips has giveji a figure of a plant which he calls Oduntnideris Lcckenhyi (Bean) Zign.," that agrees exactly with a smaJl fragment found at locality No. 19. Although the Oregon specimen is but a small fragment oi an ultimate pinna, it has several perfect pinnules that show the nerves very distinct h' . The f ( )rm ( .f the pinnules and the character of the nerves are" so entirely characteristic that a single pinnule would suffice to identify the plant. The agreement of the pinnules and nervation with those points in Phillips's figure is exact. The rachis is slender. The pinnules are remote, attached by the entire base to the sides of the rachis and slightly decurrent. They are oblong, falcate, narrowed at the ends l)y having their posterior margins cur^•e forward toward the ends of the pinna. The nervation is of the Odontopteris type. There is no mid- nerve, but the nerves go off at an oblique angle and curve suddenly and strongly away from the rachis. They are subparallel and branch several times, the center ones more copiously than the outer ones. The last branching is near the tip of the pinnule. This species differs from Ctenoptcris cycadea Sap., of the Lias, in the more falcate form of the pinnules, and especially in the more copious branching of the nerves. The fact that nearly all the branches again «See footnot. to lirynphyt., p. ,5.3. ''Geolofiy ..f Yorkshire, 3<1 e.l., p. 218. li-rn. 41, 90 MESOZOIC FLORAS OF UNITED STATES. branch near the ends of the pinnule is a noteworthy feature. The general aspect of the plant is that of a cycad, and it is by no means certain tliat it is not one." PL X\'L Fig. 1, gives the only specimen seen, and I'^ig. 2 an enlarge- ment of the same to show the details of nervation. Genus NILSONIA Brongniart. As stated under the head of Ta'niopteris, there are in the Oregon Jurassic certain tseniopteroid forms, which, although constantly without segmentation, must be placed with Xilsonia, because the lamina is inserted on the upper face of the midrib. In place of the midrilj, when the upper surface is presented uppermost, they show a cord in which the bases of the lateral nerves on each side of the midrib are inserted. When, however, the upper face is downward a midrib is visible. The plants have a rather well-defined character. The leaves were mostly thin and easily lacerated or puckered, giving sometimes a deceptive appearance of original segmen- tation. The lateral nerves are single, parallel, of equal strength from base to end, go off nearly at a right angle, and curve slightly toward the ends of the leaves. They thus contrast strongly with the true Tseniopte- rids. These unsegmented Nilsonias are the most abundant ones in the Oregon Jurassic, far surpassing the segmented forms. NiLSONIA ORIENTALIS HeCP. PI. XVI, Figs. .3-9. 1878. Nilsonia orientalis Heer: Fl. Foss. Arct., Vol. V, Pt. II (Bcitr. Foss. Fl. Sibiriens), p. IS, pi. iv, figs. .5-9. Heer has described as Nilsonia orientalis a fossil that is evidently the same as a plant that is quite widely distributed in the Oregon Jurassic. He indicates segmentation in some of his forms, l^it it may be that this is accidental and due to imperfect preservation. The Oregon specimens are all without original segmentation, but, from mutilation, sometimes show what looks like it. There are apparently in the Oregon fossils two varie- ties, one the type of the species and the other a smaller form, which may "Mr. Seward (Jur. Fl. Yorksh. Coast, p. 238) refers all these forms to Ptilozamiles {Ctenozamites) Leck- enbyi (Bean) .Natli., and regards them as cycadaceous. Thcv certainly sinudate ferns in many respects, and it is proper that they stand first as heing lowest in the grand division of seed-ljearing plants.— L. F. W, JIKASSK' FLORA OF DOFCiLAS COUNTY, ORECi. 91 1)0 regarded as a \ariety and called var. ndnur. The cliief diiiereiice is in the lateral lu^rves. Th(>re is a considorahle ditTerence in the siz(> of the ty])e forms. The largest attains a width of 4 cm. The len<;th in no case is shown, as the specimens are all fi'ajiments of leaves. The lonxtra..nlinanlv lon^. Ih.v nu.st have iK^en pc>n,h.lous. Tlu>y ran^c in wi.lth tVon. 2 nun. ..r o nun. to 10 nun Vrrv f.-w attain tl.r latter wi.ltln Most -f tl.em are r.un 5-7 nun wi.le. Tlu-ir len.^tli is unknown. U tnav l.e estimate.l hy (he fac^t that o,u> specunen was foun this species well defined, but are too tine to be seen in most cases without the help of a lens This is on account of the thick leaf substance, whu-h ditters in that respect from N. orientaUs. None of them were seen to fork. The leaf shows no trace of segmentation, and is ivmarkably free trom accidental laceration, imitating segmentation. , , ., „ , , The wider forms of this plant much resemble the fossil called by Yokovama Nilsonia ozoana;^ but the nerves are not strictly at right angles and are stronger. Besides, there is no possibility of separating them from the narrower forms. , t^ • PI XVII Fig. 1, represents the longest specimen found. It gives a good idea .'f the extreme slenderness of the leaves, for at its widest end it is onlv 7 mm. wide. It is a portion of what was a much longer leaf which probablv did not have anywhere a width much above 7 mm. The' specimens represented in Figs. 2 and 3 both occ-ur on the same rock fragment with the plant depicted in lug. 1. In Fig. 2 a fragment of medium size is represented. Fig. 3 gives a small portion of one of the narrowest leaves. Fig. 4 gives a portion of one of the commonly occurring smaller leaves, and Fig. 5 shows a portion of this enlarged. FiSchenk, Pflaiizliclie Versteinemugeu aus Richthofeu's Chiuii, Vol. IV, p. 2152, pi. liv, lig. 2li. 96 MKSOZOIC FLORAS OF rMTFD STATES. Nii.sDNi.v rTKKoi'iivi.i.oiDKs Xatlini-st Tioii Ydkoyaiiia. PI, XVHI. 187S. Xilssoiiia j)t) roj)hi/lli>i(i(s Nath. [iioii Yok."]: Foss. ¥1. vid Bjul', lift. I, p. 11 (iioimni). 1S79. Nilssoiiid pferophylloldes Nath. [noii Yok.]: <>p. cit.. lift. II, p. 72, i)L xvi. fig. 1 ; pi. xvii, figs. 2, 3. Several specimens of a plant were obtained from tlie Oregon Jurassic strata that seems to be identical with the Nilsonia pterophylloide^ of Xathorst, occurring in the Rhetic of Sweden. Xathorst's forms, how- ever, seem to l)e rather larger tliaii those from Oregon, "^'okoyama has described from Japan, with the name Dioonitcs Kotoei,'' a similar form, which, although smaller than the Oregon type, seems to be essen- tially the same and to be a Nilsonia rather than a Dioonites. It is true that Nathorst shows his leaflets as going off at an acute angle, whereas the Oregon forms have leaflets that are inserted at nearly or quite a right angle. But all of these are distorted, and the trtie angle of inser- tion may Ije tlifl'erent, or Nathorst's plants may show the leaflets of the tipper part of the leaf, while in the Oregon specimens they may belong to the lower portions, where, as is common in segmented leaves of this type, the segments make a larger angle with the axis than they do in parts nearer the ends of the leaves. Only fragments of leaves, rather poorly preserved, were obtained. Judging from the specimens, the leaflets were inserted by the entire, not widened, leases of the tipper face of the axis. They are linear in form, widest in the lower portion, and narrowing gradually toward the end. The ends in no case were preserved, so that their true nature can not be determined, and their ftill length is not known. The leaf texture seems to have been thin. In some of the specimens, owing to distortion just above the bases of the leaflets, the bases appear widened. The greatest length of leaflet seen is 65 mm., the tips not i)eing preserved. The width near the base of the widest leaflet is 5 mm. The nerves are abotit 7 in number. They are inserted at the same angle as the leaflets, are single, and of equal " Yokoyainiv in l.S'.tt niuiied a form from tlii' Mosozoic of Jiipaii .Xilssonia pteroph-ylloides as a new si)ccies (Journ. Coll. Sci. Imp. ITiiiv. .Japan, Vol, VII, Pt, 111, 1894, pp. 207, 22S. pi, xxii, (ig.s, S-IO; pi, xxv, fig, 7). It prove.s to be the XiUoiiid adijornirn Kont, of (lie Shasta group. See ]), 2.52, — L, F, W. ti Jouni, Coll, Sii. Imp. Univ. .Japan, Vol. Ill, Pt. I, pp. 44-45, pi. vii, figs, labc, le; pi, xiv, lig, 14, jrUASSIC FLORA OF l)OL'(iLA8 COUNTY, ORPX). 97 strength fioin near their l):iso.s to their ends. They are distinctly defined, l)Ul slender. PI. XVITT. Fiii. 1, shows the largest specimen seen, in which the leaflets are considerably distorted, none being entire. The nan-owing above the bases of the leaflets, due to distortion, causes the bases to appear too wide. Fig. 2 gives leaflets of the smallest size, they being shown on onl.\- one side of the rachis. Fig. o shows foui' of the seg- ments enlarged. I'ig. 4 gives a fragment of a leaf with the longest leaflets seen. l'"ig. - abundant at some of the localities and the large number of specimens obtained give a good idea of its charactei'. The leaf was evidently quite long in propor- tion to its width, but in all cases it was small. Most of the Oregon forms are rather smaller than the two given by Schenk, but specimens can be selected that equal these in size. The leaves were linear elliptical in shape, wider in the central portion, and, fi'om the shortening of the segments, narrowing toward the base and apex. The longest fragment seen is 85 mm. long, without showing the base and tip. In this the longest segments are in the upper end, showing that this was near the middle of the leaf. These segments are about 7 mm. long, so that the entire leaf was aljout 17 cm. long and 15 mm. wide in its middle portion. The leaves apparently had a petiole. The leaf substance was pretty MON xi.viii — 0.5 -7 98 ME.SOZUIC FLOKAS OF UMTED STATES. thick, for it leaves a film of coal, ;ui(l the lateral nerves of the segments are usnall>- hidden. The axis of the leaf is quite strong and rigid in proportion to the dimensions of the .segments or leaflets. The leaflets on different leaves in the same positions vary a good deal in length and width and in the proportion of length to width. They are, as stated, alwaj's shorter toward the ends of the leaves, and toward the base are often proportionally wider. Sometimes in that position they are markedly irregular in form, the width being abnormally great. In descriliing the leaflets we must take the average or more common forms found toward the central portions of the leaves. The leaflets are at- tached nearly or quite at a right angle, by their entire t)ases, which are of the same width as the main portion of the leaflet. They ai'e linear with parallel sides. The attachment is to the sides of the midrilj or axis of the leaf. They are slightly falcate, l)ut may be straight. The tip is oljtuse and rounded ofl', but narrowed somewhat by the pos- terior margin curving foi"w^ard. The nerves of the leaflets are few in number, 5-7, and slender, being generally not visible without the help of a lens. They are attached at the same angle as the leaflets, and are parallel to one another and to the margins of the leaflets. They are not forked, and are of equal strength from their insertions to their ends. The average leaflets are about 8 mm. long and 2 mm. wide. They may, however, be only 3 mm. long and 1 mm. wide. The plant has in the form of the leaflets a resemblance to the two, fossils figured by Lindley and Hutton as PterophyUum pecten (Phill.) L. & H.° and Zamia taxina L. & H.** It resembles the latter more closely, but the leaflets are mostly smaller and closer. It has the leaflets much closer than the former. PI. XIX, Fig. 1, gives the largest fragment seen. It shows well the increase in the length of the leaflets in ascending from the l)ase, and tlie greater width of the basal ones. Fig. 2 gives a fragment of the middle portion of a leaf having average leaflets, and Fig. 3 shows a portion of this enlarged. Fig. 4 represents a fragment from a similar part of the leaf, with narrower leaflets. Fig. 5 gives a basal portion of a small leaf. Fig. 6 shows a fragment of the middle part of a leaf, with wider and shorter leaflets than usual. In these the nerves are rather more remote than is common. Thev are 5 in inuiil)er. oFoss. Fl. Gt. Brit., Vol. II, pp. 61-62, pi. cii. 6 0p. cit., Vol. Ill, p. 67, pi. clxxv. .HKASSIC FL01{.\ OF DOUdLAS COrNTY. OHKd. 99 Tho ])l:uil isv(M-\- ;il)Uii(laiit nt localilx" No. 7. aiul is lioro only iiifoi-ioi' in (luantity to I'li rojtliiiUuni rnjnuihalcHsc. (Sec ]). Id'-'.) ll i>^ voiy aluuidaiil at hicalily No. 4 and is coniinoii al localilx' No. 2. It occurs at localities Xos. S, 12, and H). PTEHoriivi.LrM <'o\ rua I M Schciik. I>l. XIX, Ki^'s. 7 11. 18S3. rhriijilnillinii i-oiitiiiiiiDii Schciik: Plhinliclic N'cr.stcincnini^'cii aiis Kiclit- liolVn's Cliina, \'ul. I\'. |). --'(i-'. |)l. liii. li.L'. •>. The fossil from tlie Jurassic of China, described hy Schenk i\i> Ftcro- phylluni contignum, seems to be the same as a ])lant that is not rare at some of the Ore.son localities. Schenk seems to have had only one specimen and could not determine its full ch.aracKM-. 'V\\v Oregon specimens ai'e fragments mostly of the middle portions of leaves. The midril) is very strong and rigid in proportion to the size of the leaflets. The leaflets are quite thin in texture. They are lineai- in form, of the same width from base to apex; not falcate, but straight. They are very obtuse at their ends and symmetrically rounded off there. They stand at right angles to the rachis. Their form is much like that of the leaflets of /■•. (rqualc, but. they are mostly narrower. They vary a good deal in size, being in some cases almost as small as the smallest of the leaflets of P. NnUwrsti. The average leaflet is about 2 cm. long and a little more than 2 mm. wide. The smallest forms are somewhat less than 10 mm. long and 2 nmi. wide. These are rare. Schenk shows the leaflets as simply touching one another, but in many of the Oregon forms they not only touch but seem to be united, so as to show l)etween two adjacent leaflets what looks like a strong nerve or cord. This may be simply a line of the rock matter scjueezed l)etween the adjacent leaflets. The distinguishing character of the plant is the constant close approximation of the leaflets. The nerves are slender, and to be seen require the help of a lens. They are al)Out 5 in number, simple, and ])ai'all(>l to one another and to the margins of the leaflets. PI. XL\, Mg. 7, gives a portion of the middle part of a leaf carrying leaflets of the largest size, and Fig. 8 shows a poi'tion of this enlarged. Fig. 9 shows a similar part of a leaf witli leaflets of the narrowest kind, and Fig. 10 a portion of the same enlargetl. iMg. 1 1 represents a poorly pre- served portion of a leaf with the narrowest leaflets. The plant is most common at locality No. 7, where it is abundant. It is also found at localities Nos. 4 and 19. 100 MESOZOir FLORAS OF UNITED STATES. PxEKOPiiYLLrM ^QUALE (Broiiojiiiart) Nathorst. PI. XX. IS?."). Xilsonin ^ spqvnJis Rronijii.: Ann. Sci. \at. Pari.s. \'<>l. TV, p. 219. ])I. xii, fig. 6. 1828. Pterophylluhi iJuhnun Bi-ongn.: Prodrome, p. 95. 18.38. Zamifes xqunlis (Bron^n.) Presl in Sternberg: Flora der Vorwelt. Vol. II, p. 198. 1S41. PtUophyllum tequale (Brongn.) Morr.: Ann. & Mag. Nat. Hist., Ser. I. Vol. VII, p. 117. 1878. PffmpltiiUum ^equale (Brongn.) Nath.: Foss. Fl. vid. Bjuf. Hft. I, p. 11 (nomen); K. Svensk. Vet.-Akad. Handl., Vol. XVI, No. 7. pp. IS, 4S, pi. ii, fig. 1.3; pi. vi, figs. 8-11. Xathorst has given from the Rhetic of Sweden, with the name Ptero- pln/llum cequale" two forms that seem to me to l^elong to different species. The plant figiu'ed in his Floran vid Bjuf, Heft II, pi. xv, fig. 11a, as P. (vquale rectangnlare seems to be identical with P. rajmahaJense '' from India. This I infer from the fact that the tnie P. ccquale and P. rajmahalense are both abundant in the Oregon flora, and are quite constant in character, with differences sufficient to denote that they are different plants. It is true that a few forms are intermediate between the two, but they are exceptions, and not sufficient to establish a passage of one form into the other. In plants of this type there could not be a variation in the size of the leaflets without approaching the one or the other type, and in the great numljer of specimens that were obtained they are to be expected. They should be interpreted not as true passage forms, but rather as aberrant "and accidental. It is possible to select in the many specimens a few forms that would make a passage from the largest of the mpnahalense type to the most minute of the P. Naihorsti leaves. The leaf substance of P. (vquale is thinner than that of P. rajmahalense. The rachis is comparatively slender. No entire leaves were seen. The leaflets are long in proportion to their width, appearing slender, and con- trasting in that point with those of P. rajmahalense. They are linear, with the margins parallel to each other, and are mostly slightly falcate. "Flciraii vid Ilnfraiiiis, p. 18, 1>I. ii. li^. 1^5. pp. 4.S 4!). pi. vi, lig-s. S-11: Fliirmi viil Bjuf. pp. (i7- mi(ldl(> portion of the leaves may he taken as having a length of IS mm. and a width of .5 mm. Those of maximum size from the same part of the leaf may attain a length of 20 nun. and a width of 8 mm. The smallest leaflets from similai' parts are lo mm. long and 4 nun. wide. The nerves are very fine and raiely visil)le. They can he seen distinctly only with the lielp of a lens. Tlu>y are attached at a right angle to the axis, are parallel to one another, and not forked. They vary somewhat in number according to the width of the leaf, .\bout 15 may be taken as the average, l)ut 17 or more may occur. They do not vary in strength from one end to the other. Most of the forms given l)y Oldham and Morris fi'om the Indian Rajmahal series are larger than the average forms from Oregon, but the Rajmahal Pterophylla generally show a remarkable robustness. This plant was found in the Oroville flora" in a few specimens forming an important connecting link between that and the Oregon Jurassic. Heer describes from the Jurassic of Siberia a fossil that he names Ptero- lilniUiun Sensinovinnnni.'' This seems to be the same as the plant now in (juestion. He separates it from P. rajmahalense because it has fewer nerves, but he says the nerves are obsolete, and, as he had in hand only one small fragment, it is not probable that he saw all the nerves. In the hundreds of specimens fountl in Oregon only a few show the nerves distinctly enough to count them. PI. XXI, Fig. 1, shows the lower portion 'of a leaf with leaflets of medium size. Fig. 2 is an enlarged portion of Fig. 1. Fig. 3 gives a similar part of another leaf, with short, wide leaflets, the lowest ones abnormally wide, and also a portion of the petiole. Fig. 4 gives the ter- minal part of one of the smallest leaves, showing the two end leaflets abruptly much reduced in size. These are enlarged in I'ig. 5. Fig. 6 gives the basal part of a leaf carrying leaflets of the greatest width, and showing an abnormal widening of the liasal ones. Fig. 7 shows nearly an entire leaf with leaflets of the largest size. aTwentieth Aim. Kep. U. S. Geol. Surv.. I't. II. 1900. j). 3o4, pi. Ivi, figs. 4, .5. ''Fl. Foss. Arct., Vol. IV, Pt. II, p. 10.5, pi. .\.\iv, fig. S. 104 MESOZOIC FLORAS OK UMTED STATES. The plant occui's with an iiniiioiise number of imprints at locality Xo. 7. where it stands next to the (iinkgos in miml)er. It is found also at localities Xos. 1, 4, 14, and 19. Pterophylli'm MINI'S Broiigiiiart ? PI. XXI. P^ijjs. s. 1). 1825. I'ffrophi/lluin minus Brongn.; Ann. Sci. Xat. Paris, Vol. IV, p. L'lO. pi. xii, fig. S.'^ Several specimens of a small plant that is much like the /^/r/v)/)/(;///(//;i minus figured hy Lindley and Hutton'' are foimd at locality Xo. 7. The plant is somewhat smaller than the form figured in Fossil flora, hut is of the same type. The leaflets are about 5 mm. long and 3 mm. wide. They are closely placed, touching by their edges. They are at right angles to the midrib and of equal width from l)ase to tip. The ends are truncate or slightly rounded. The nerves are about 12 in number, single, perpendicular to the midrib, and parallel to one another. They are slender and can be seen only obscurely, even with a lens. The plant sometimes approaches the wdder forms of PtcrophyUum NathorsH. It may be a X^ilsonia, but a strong midrib is always shown. The amount of material is too small and too poorlv preserved to permit positive identification, with the plant of Lindley and Hutton. PI. XXI, Fig. 8 shows the specimen natural size and Fig. 9 the upper part enlarged. "Professor Fontaine does not refer to this figure nor cite this memoir, and Mr. Seward also ignores it. It is an obscure and little-known paper, but important as being the one in which the genera Pterophyllum and Nilsonia were first named. The plates of tlic early volumes of the Annales are difficult to find, being in quarto form and u.s\iall_v bound up in atlases that cover several volumes of the text. They are wanting in munv libraries and are generally overlooked by bibliographers. The plants were from the lihetic of Hi)r in Scania, but Lindley and Hutton identified a Yorkshire Oolitic form with this species, and it is their figure that Professor Fontaine refers to. Lindley and Hutton give the name Pterophylhim Nilsoni to another figure on the same plate, identifying it with the Aapleniopteris Nilsoni 'i figured by Phillips in his Geology of York- shire, 1829, pi. viii, fig. .5, which in turn was supposed by him to be probably the plant so named by Stern- berg in his Flora der Vorwelt, Vol. I (Tentainen, p. xxii, also index and index i basal portion, perpendicular to the tixis. In the higher i)arts they become more and more inclined. They were usually somewhat falcate, more so in the upper parts of the leaf. The leaf sul)stance was thick aiifl dense. They are slightly expanded where the base is attached, and keep their width unchanged to above the middle of the leaflet, then narrow gradually to the t ip. They are very nttrrow and litiear in form. The tips are obtusely rounded off. In the same position on the leaf they may vary somewhat in width, a few being a little wider or narrower. Taking the leaflets from the middle portions of the leaves for measurement, there is some small variation in different leaves. The widest leaflets seen did not have their ends preserved. They have a width of 3 mm. The narrowest leaflets are only 2 nun. wide. The longest leaflet seen, probalily not actually the longest, is fi cm. long and 2 mm. wide near the base. The leaflets are rather remotely placed. The nerves are slender, and not easily seen without the help of a lens. They are unbranched, attached at the same angle as the leaflets, and about 5 in number. 106 IVIESOZOK' FLORAS OK rXITED STATES. PI. XXII, Fi^. 1 shows a fofiii with the widesi leaflets; Fig. 2 gives a fragment showing the greatest lengtli of leaflets in leaves of average width, and Fig. 8 a few of these enlarged a little: Fig. 4 shows a pretty large fragment witli leaflets of the narrowest kind; Fig. o includes two fragments with leaflets of average width, having some irregular in size. The plant is quite common at locality Xo. 2, and occurs also at locality Xo. 1. Ctenopiivi.hm rACHYXERVE Foiitaiiic n. sp. PI. XXIII, Figs. 1-4. A plant regarded as a new species of Ctenophyllum was found, with several specimens, in the Oregon flora. It has some resemblance to Ctenophyllum gninclifolium Storrsii," hut is a much smaller plant. The variety Storrsii, in the description of the Oroville flora, was erroneously drawn, with anastomosing nerves. The species now in question is rare, and the specimens are mostly poorly preserved, but it has such a dis- tinct character that it differs certainly from any previously described form. The midnerve or axis of the leaves is not well shown, but seems to be slender. The leaflets vary somewhat according, apparently, to their position on the leaves. They are attached by their entire base to the sides of the axis. Apparently they stand at right angles to the axis in the lower portion of the leaves and are there straight. In the upper parts of the leaves they are inclined to the axis at angles more acute as the ends of the leaves are approached. In these positions they are slightly falcate. All are linear in form, widest in the lower part, and narrow toward their tips, having obtuse ends. In some of the specimens distortion has caused the insertion of the leaflets to appear somewhat widened, which is not really the case. The upper leaflets are more closely placed than the lower ones. The latter are sometimes quite remote. The upper falcate leaflets are the only ones showing their full length. They are 5 cm. long and 4 mm. wide. The leaf substance was thick and leather-like. The most characteristic feature is seen in the nerves of the leaflets. These are 5-7 in number, attached under the same angle as the leaflets, mostly single and parallel to one another. They are peculiarly strong and stand out like threads, causing the leaflets to appear striated. In the upper falcate leaflets no n Twcntipfti .\nn. Rep. V. S. Opol. Surv., Pt. 11, 1900, p. 3.59, pi. liii. fig. .3: pi. Ixii: pi. Ixiii, fi^. 1 : pi. Ixvi, fig. 3. JUKASSIC FLORA OK DOlCiLAS COUNTY. OKE(;. 107 forkiiifi was seen in the nerves; in the lower ones a nerve oceasionally forks ai)ove the point of insertion. IM. Will, I'i^. 1, gives a consichM'ahlc poi't ion of llie upper pai-t of a leaf with ol>rK|uely place:. 2 represents a fragment of tlie basal pari of a leaf containing the basal portion of some remote, nari'ow leaflets. Fig. :i shows a similai- pai't of a leaf carrying the basal parts of several larger and more closely placed leaflets, fig. 4 shows two of these leaflets enlarged. The plant is rare eveiywhere. It is most conmion at localities Nos. IS and 19. CtENoI'IIVLU'M WaK'DO Fcillt:iill('. PI. XXIIl. Fi-s. .-, 1_>. iSDf). CUnophjnumVt'drd'ii Font.: Am. -louni. Sci., Itli scr.. Vol. II, p. --'74 (iioiihmi). 1<)()(). Ctenophylhii,, Wnnlil Font.: Twcntirlli .Vmi. \U•\^. V. S. (icol. Surv., 1S'.)S !((). Pt. II. p. ;!")7. pi. li-\; pi. Ix; 1>1- 1-^vii, lig. :>. A considerable numi)er of specimens of a plant that agrees well with Ctenophyllum Wnnlii were found at some of the Oregon localities. The>- are ver>' fragmentary, and do not compare with the fine specimens found in the Oroville flora. Nearly all the specimens are fragments of unat- tached leaflets, pooi-l\- preserved. The midrib indicated on the Oregon plants seems wider than that seen in the Oroville specimens, for one shows a width of 5 mm. The leaflets have about the same range in width as those of Oroville. The widest have a width of about 2 cm. and the nar- rowest aljout I cm. The nerves are numerous, fine, and closely placed. Thev are prominent, and not immersed in the leaf substance, but stand out like threads. The\- fork mostly at their base, and sometimes, but rarely, hio-her up. Some are single. Very rarely an anastomosis may be detected, l)ut it is clearly not an essential feature. I'ig. o shows parts of several leaflets evidently once united to a com- mon midrib, and Figs. 6 and 7 are enlargements of two of these. Fig. 8 includes portions of two lai'ge leaflets, probably of their middle parts. V\g. 9 gives part of a small leaflet. Fig. 10 shows part of a midrib and several attached fragments of leaflets shown on only one side of the midiib. Fig. 1 1 shows a fragment of the upper part of a leaf with portions of several leaflets attached. Fig. 12 is an enlargement of Fig. 1 1 . It occurs at localities Nos. 1, 2, 7, 11, 17, and 19. 108 MESOZOIC FLORAS OF UNFrED STATES. Genus PODOZ.AJMITES Friedrich Braun. PODOZAMITES PCLCHELLUS Heel'. PI. XXIV, Figs. 1-10. 1876. Podozamitefi pulchi]his Heer: Fl. Foss. Arct., Vol. IV, Pt. I (Beitr. z. Foss. Fl. S])itzbergens), j). 3S, pi. ix, figs. 10, Ihi, 12b, 13, 14. 1S76. Zi Ward n. roinl)." I'l. XXV. Fii;-. '•: I'l. XX\I. IS'i.S. Zam'm Inncfilhliu Bnniu:!!. ; I'rodroinc. j)]). '.M. I'.t'.» OionuMi). 1829. Cyraditeti su/ricdiills I'liill.: (i(>()l()<,ry of Vnrksliiiv. ])]). 1 4s. is'.i. j)!. vii. lig. L'l. 1,834. rttni\faJratii L. & II.: Foss. Fl. (it. I'.iii.. \nl. II, |.. (i:;. pi. nm\i\\.) Morr.: Ann. cV^ Ma^. Xat. Ili.st., 1st Scr., Vol. VII, p. 116. IStU. I'liropliiiUninfcilaifiiii, (L. .<: II. I San(ll)cr>;cr [non Xatli.]: ' NCrli. d. Xaliirw. \'(>r. in Karlsrnlii', Ilrl't I, \). :!•") [d]- A inmilK'r of spociincns of thi.s Hue plant woi'o obtaiiipcl. Unfortu- nately the stratum whicli contains most of them has no cleavage and tends to l)reak across th(> plane of the leaves. Hence the specimens procm-ed are smaller than the parts contained in the rock. The figures given hy Zigno' of this plant very accurately represent it as found in the Oregon strata. 1 do not find, however, on the axes, the regular reticulation given in Zigno's Fig. la. The marking on the axes is an irregular wrinkling or puckering of the epidermis, seen onl>- when that is present. The axes seem to be very robust, but the amount of vas- cular tissue is not so great as might l)e supposed from its width. They were apparently succulent, with a large proportion of cellular tissue. The apparent width is increased also by the continuation, over th(> axis, of the thick epidermis of the base of the leaflets. The leaves must « The name Ctenis falcata L. & H. certainly can not stand. Lindley and Button themselves state that it is the Chjcmliteii sulcicaulin of Phillips, of which that author gives a fair figure in his Geology of Yorkshire ( 1829) . That Lindley and Hutton had better material and made a better figure is no reason for changing a name, as this would permit anyone at any time to make a new name if better material were discovered. The name Zamia Inmj'ifoUa of Brongniart, given by Mr. Seward (Jur. Fl. Yorksh. Coast, p. 23.5) may be ignored as a nomen imdum, no description or figure of it having ever been published either under that name or under the name Zamiles lo7i^ifoliiis, first used by Morris in 1841, and later by Brongniart himself in his Tableau (l)p. B2, lOfi). The fact that Mr. Seward found in the Paris Museum a specimen labeled Zamia longifolia l)cl()nging to this species is scarcely sufTicient to justify giving this specific name to the plant. If it had been the I vpe, so designated by Brongniart in his Prodrome, or even on the label, it might have been accepted under some codes, as, for example, that of the Ornithologists' Union, Canon XLIII, p. .53; but Mr. Seward does not sav that the label wa.'i in Brongniart 's handwriting, and, moreover, the specimen was found at Caylon near Scarborough, while Brongniart (Prodrome, p. 199) gives Whitby as the locality. All things considered, tlicre- fore, it does not seem possible to adopt Brongniart's name, and it must become a synonym.— L. F. W. '' This combination is usually credited to Sehimper (Pal. Veg., Vol. II, p. 137), who does not meniion Sandberger's paper. The Pterophijllum :> falcatum Nath., Fl. v. Bjuf, Hft. 11, p. 71, pi. xiii. figs, lli, 17 (1879), wliatever it may be, is a preoccupied name and must be cliangcd. — L. F. W . cFlor. Foss. Form. Oolith., \' anastomose rather rarely and irregularly, ])y sending off a branch which unites with an adjacent nerve at a very acute angle. PI. XXVn, Fig. I, gives the most complete specimen found. Fig. 2 shows one of the leaflets enlarged. Fig. 3 shows th(> base of a leafl(>t, indicating its mode of attachment. This is slightly enlarged in Fig. 4. Fig. 5 gives what seems to have been the upper part of a leaf. It lias the basal parts of several leaflets attached to the midrib, but the attach- ment is oliscured by distortion. PI. XXVIII, Fig. 1, shows the basal parts of several leaflets, so placed as to indicate a former attachment to 116 MESOZOIC FLORAS OF LMTFL) STATES. a common midi'il). It is not certain that these belong to the plant in question. If they do they belong to the upper part of a leaf. The plant is not common. It is most ai)un(lant at locality No. 7. Several specimens were found at locality' Xo. 2. Ctenis grandifolia Fontaine PI. XXVIII, Fio:s. L>-8. 1896. Ctenis grandifalhi Font.: Am. .Toiiiii. Sci., 4th ser., \'iil. II, ]>. 274 (nonicii.). 1900. Ctemis grandifoUa Font.: Twentieth Ann. Rej). V. S. (ieol. Surv.. 1S9S-99, Pt. II, p. ."^54, pi. hii, fig. 2; ph Ivi, figs. B, 7: pi. Ivii. Several specimens of a plant much like Ctenis (jrdndifolia were obtained from the Oroville localities. This plant was foimd in the Oroville flora. The Oregon specimens do not add anything to the character made out from the Oroville plants. They are much too dis- torted and mutilated. Only fragments were found. From these neither the shape nor the size of the leaflets can be determined. The attach- ment seems to have lieen liy the whole of a somewhat ex])anded l)ase, which in one specimen seems to ])v deciu'rent, but this is probably due to distortion. The leaflets in the Oregon specimens apparently some- times had great width, equaling 5 cm. The narrowest had near the base a width of about 25 mm. The nerves are cjuite remote, strong, and sharply distinct, Imt they are immersed in the leaf sulistance. They are approximately parallel and anastoniose rather freely at long inter- vals, in the same manner as those of Ctenis orerviUensis. PI. XXVIII, Fig. 2, shows the most complete specimen, but it is greatly distorted. A portion of the midrib remains, and to this the three leaflets still preserved were formerly attached. But the attached parts are not now visible. The leaflets indicate a width of 5 cm. They are doubled upon themselves along the middle line of their length, owing to being crushed clown into the rock. Fig. 3 gives the l)asal part of a fragment of a wide leaflet, showing nerves rather vaguely. A small portion of this is shown enlarged in Fig. 4. Fig. 5 represents a fragment of the narrowest form of leaflet with the ])ase preserved and showing the mode of attachment, but probably distorted so as to cause an apparent decvuTence. I*'ig. (> shows the basal portion of this JURASSIC; FLOHA DOUCJLAS COUNTY, ORE(}. 117 enlarged. V'l^^. 7 and 8 show a frafinient 4 em. wide, indicating the nerves quite well. The plant is not abundant. It is most coimnon at locality .\o. 7. It occnrs moi'c rai'ely at localities Xos. 2, fl, and IS. Ctenis AiHicri-ATA l-"i)ntiiinc? I'l. XXl.X, Ki^^ 1. IS!)(). CttniK (lunniliila Font.: Am. .Idurii. ScL. 4lli scr.. Vol. II, \^. I'Tl (iioincii.). 10(10. Cliiiin iniri<-iiliit(i Fiint.; 'twcnticl li Ann. I\i'|i. f. .S. (icol. Surv., ISilS !l!l, Pt. II, |). .".."><'., 1)1. Iviii, lii^s. 1 :;. At locality No. 7 a single .specimen of a lai'gc Ctenis was ohtained that agrees qnit<» well with (\ auricniata, a plant first found in the Oroville flora. Tiie preservation is not perfect enough to make the identification positive. A portion of the midi'it) was obtained having on one side the basal i)arts of two leaflets, one pi-etty well preserved. This is .") cm. long and 'A cm. wide at base. It is attached apparently l)y all of an unnarrowed base, whi(;h is not auriculate. The leaf sul)stance was very thin and left very little coal on tlie i-ock. The nerves are strong, very remote, single at theii' bas(\s, but anastomo.se freely about 15 mm. above tluai' ba.ses — more freel\- than in the Oroville specimens. It may t)e a new species." Genus ENOEPHALARTOPSIS Fontaine. Encephalartopsis * OKECiONENsis Fontaine n. .sp. PI. XXIX, Figs. 2, -.i. A single fragment of a leaflet, with resemblance to those of Encephalartos, was found at locality No. 4. It is of doubtful affinity and is mentioned because it certainly differs fi'om th(> other cycads of the Oregon flora. It is narrowly linear, narrowing gi'adually toward its end. It has a thick, dense texture, which gives it a very rigid aspect. Apparently it had a firm, almost hoiiiy epidermis. It must have had a considerable length, foi' the part preserved, though a small portion of the original leaf, still shows a length of (\ cm., with a maximum width of 4 mm. It had spiny points going oh' at long intervals and tin-ning " Tilt' spcriiiiiMi was oil two pieces of rock, one the counterpart of one-lialf of the leaf shown on the other. In the latter tlie leaf was foldeil on lioth sides of a thin shili, the nieliis running along the thin edge. The figure (PI. XXIX, Fig. 1 ) shows the frond unfolded and givi's u fair idea of it.s nature. — L. F. W. 1 1 ^^ MESOZOIC FLORAS OF UNITED STATES. ■""" strongly awa>- from the leaflets. These are now j)reserve(l only in their lower portions. The nerves are single, strong, parallel, and vaguely shown. (Iciius ('Y('Al)EOSPERMOI Saporta. Cyc ADKOSPKHMIM OREGONENSE FolUaillc II. S|). PI. XXIX. Fij;. 4. A single specimen of a small nut-like seed was found at locality Xo. 2. It seems to be the seed of some cycad. It is hard and ])ony, with a smooth polished surface. It is elliptical in form and narrows to both ends, one of which is rounded off, while the other is somewhat prolonged into a short acute tip. It is widest in the middle and is there 6 mm. w'ide. The length is 10 nun. 't->' CyCADEOSPEK.MI'M OVATl'.M P'oIltailK' 11. sp. PI. XXIX, Fill. 5. A single specimen of a sniall nut was obtained from locality No. 7. It is apparently the seed of some cycad. It has a hard, l)ony, and smooth surface. It is widest at its base and is rounded off there. It narrows to the opposite end, which is obtuse. At this end there is a ridge which runs up a little way into the body of the nut. This may l)e due to puckering from pressure, causing a wrinkle, but this is not pi-ol;able. The nut is 10 mm. long and 6 mm. wide at base. Order BENNETTITALES. Family BENNETTITACE.E. Genus WILLLUISONIA Carriitlu-rs. Wii.LiAMS(»Ni.v oREGONENSis Fontaine n. sp. PI. XXIX, Fig. 6. A single fine specimen of the inflorescence of a Williamsonia was found at locality No. 7. There is nothing to show a connection between this plant and the numerous cycads found at that locality. On the same piece of rock, however, a well-preserved fragment of Pterophyllum cequale (Brongn.) Nath. occurs. The fossil is large and globular in form, showing a whorl of bracts incurved so as to hide any object wdthin them. The ijracts are thick in texture and without any midnerve. JTRASSIC FLORA OF DOl'OLAS COINTV. OHKC. Ill) linear in foiin witli obtuse ends. The glol)ular infioi-escciiee is l)ro;ider than lonmbling a leaf was obtained from locality No. 7. It is a good deal distorted and has at present an ovate shape. It is somewhat convex from pressure, and on casuid inspection looks somewhat like a nut. It was, however, too soft for this to be the case. It is liroadest at base and tapers ovately to the tip, which is obtuse. The base shows traces of an original cordate shape. At about the center of the basal margin there is a roiuided depression that looks as if the object had been joined to its support by articulation. The fossil looks more like the bi'act of th(^ involucre of some form of Williamsonia. WiLLiAMsoNiA ? sp. Fontaine. Bracts of Williamsonia ? No. 2 {a, h, c). PI. XXIX, Figs. S 12. Three objects were found at locality No. 19 that at first sight might seem to ))e nut-like seeds. They are, however, most probably thick, fleshy bracts or leaves, and appear to belong to the inflorescence of some Williamsonia. Although, as shown on the rock, they are slightly con- vex, they are not enough so for a nut-like seed, and l^esides they show evidence of having been too soft for such objects. The leaf substance is represented by a film of coal of considerable thickness, which, how- ever, is not always present, having l)een stripped off. The slate under this coal film shows a distinct nervation, which is not visible on the 120 MKSOZOIC FLOK'AS OF rMTED STATES. leaf svibstance. This indicates that the nerves are disclosed only on the lower snrface. The nervation resembles that of Xeuropteris of the Carbonifei'ous, but tliere is no niidnerve visible, although it may be present, as the nervation is not fully disclosed. The nerves appear to radiate from the Ijase of the fossil, and they fork several times. The surface of the coal film, which represents the supposed l:)ract, is granu- lated witli ininut(> papilla' of no definite form. They may l)e due to the shrinking of tlie leaf. The general aspect of these fossils suggests that they may be l)racts of the involucre of some ^^'illiamsonia. Tlie three found are apparently essentially the same, but present some unim- portant differences in form, which are shown in the figures given. They are distinguished as a, h, c. Bract a is elliptical in form and not symmetrical, one lateral margin being more strongly convex than the other. It tapers gradually to a sul)acute tip, which is well shown. The ))ase is imperfect. I^ract b is of the same general form, but is broader and propt)rtionally shorter. The basal portion is wanting. Bract c is ovate elliptical and nearly symmetrical in form. PI. XXIX, Fig. 8, represents bract a, enlarged in Fig. 9; Fig. 10 bract b, and Fig. 11 bract c, enlarged in Fig. 12. Order GINKGOALES. Family (aNKGOACE.F]. Genus GINKGO Ka.Mupfer. The Ginkgos of the Oregon Jurassic are the most important fossils in it. This is due to the fact that they show a great mmil)er of speci- mens and a very considerable variety of forms, which are quite well preserved. The leaves are preserved mostly entire, and the condition in which they may he ol)tained depends on the splitting of tlie rock containing them. Unfortunately the cleavage of this is generally poor, othei-wise very perfect specimens might be ol)tained in greater num- bers. At some of the localities the impi'ints of Ginkgo leaves are exceedingly numerous and may be obtained by the hundreds. The faces of the rock, when exposed by splitting it, are covered with numer- ous impressions, often of diffeictil character or type, while the variation in the leaves is considerable. 1 am not sure that they are not all modi- JURASSIC FLORA OF DOl'OLAS COUNTY, OREG. 121 ficatioiis of Ihc ratlioi' polymorphous species Ginkgo (lifiitata (Bronsn.) Heer. The hmits of variation in the same species, for leaves of the Ginkgo type, are not fully known. Mr. Seward and Miss Gowan, in their paper on diiikno hiloha:' have shown that the living Ginkgo ti-ee is capable of considerable variation in its leaves, and that some of these forms even r'eseinbl(> Jin-assic types. While foi-ms of the leaves of the same sjiecies may lie s(>le('ted that would show such a variation, the significance as fossils of ihi-se variant forms would be better understood if it were known what proi)oilion in lumibers they b(>ar to the normal forms and in what number they would occur in a foi'tuitous collection of leaves produced by a fall from the same tr(M\ In the Oregon localities many portions of I'ock ar(> covered with leaves of (Ufferent types in about equal proportions, indicating that each type belonged to different trees, on which it was the predominant foi'in of leaf. The great(>r numbei' of the Oregon Ginkgo leaves may lie brought under several types of previously descril)ed forms. Rut these may l)e connected by intermediate forms, which occur in con- sideral)le numbers. Most of the previously described types illustrated in the Oregon (Jinkgos are those given b>- Heer for th(> Jvu'assic of Siberia. It will perhaps give a better idea of the Oregon forms if they are described as belonging to the previously described species, which they most resemble, treating under separate heads those that can not be thus grouped. UixKuo DiGiTATA (Brunjiiiiart) Heer. Pi. XXX, Vlfr^. 1-7. l,S-2<). Sphirnnpteris latifoVui Pliill. [noii Broiign.] '' Geology of Yorkshire, pp. 148, 1S9, pi. vii, fig. IN. 18.S0. ('iicJopteris (ligitatd Broiign.: Hi.st. Veg. Fos.s., Vol. 1, p. 21!l, jil. Ixi l)is, figs. ■_', :]. l,s:{(). Adianiites digltatus (Broiigii.) Gopi).: Sy.st. Fil. Foss., p. 217. ],S4:i. Haiera digitata (Brongn.) Fr. Br. in Minister: Beitr. z. Petrefacten-Kunde, Vok II, Heft VT, ]). I'l. 1865. Cyclopteris incisa Eiehw.: Letha-a Rossi('a, Vok II, p. 13, pi. iv, lig. G. " T\w maitlpiiliuir tree, l)y A. C. Seward uiul Miss .1 . ( Iciwun : Annals of Botany, Vol. XTV, No. LIIT. ^tarcli , 19()0, pp. 109-1.54, pi. vili x. '' Broiijjniart 's Carboniferous speeies Sphnwpleri.f Inlifnlia dates from tlie Prodrome (p. .'il ), one year earlier than tliis, whieli is tliercfore to be dropped, although antedating all other names for this plant.— L. F. W. 122 iMESOZOlC FLORAS OF rNFrFD STATES. 1874. Ginl-go digitata (Brongn.) Heer: Rcgel'.s (iaitenflmu. .laliirr. XXIII, p. 261, pi. dcccvii. fig.s. 1-4. 1874. Ginl-go integriuscula Heer: Kegel's Gartenfloia, -lahrg. XXIII, j). I'Ol, ])1. dcccvii, fig. 5. 1876. Ginl-go dig'ttnta (Brongn.) Heer: Fl. Foss. Arct.. Vol. TV. Pt. T (Bcitr. z. Foss. F'l. Spitzbergens), p. 40, pi. x, figs. 1, na h, (i. 1876. Ginkgo digitata biloba Heer: Op. cit., p. 41, pi. viii, figs, la, laa. 1876. Ginkgo digitata quadriloba Heer: Op. cit., p. 42, pi. x, figs. ;->», h. 1S76. Gitilgo digitata multiloha Heer: Op. cit., p. 42, pi. x, fig. 2. 1876. Ginkgo digitata angustiloia Heer: Op. cit., p. 43, pi. x, fig. 4. 1876. Ginl-go integriuscula Heer: Op. cit.. p. 44, ])1. x, figs. 7-9. 1878. Salishuria digitata (Brongn.) Saj).: Plantes Jura.ssiqiies. Vol. Ill, p. 294, pi. clx [xxxii], figs. 1-5. 1881. Ginkgo digitata integriuscula (Heer) Kollbrunncr: .Tahresb. d. Ostschweiz- Geogr.-Comm. Ges. in St. Gallen, 1880-1881, pp. 62, 77." I will place with this species those leaves that have the following character, ignoring possible sporadic variation: Leaf substance thin; leaves small, with a wide spread, being wider than long; segments numerous and not deeply incised, wide in proportion to their length, closely placed and widest near or at their ends, so that they have an ol:)Cuneate form; ends obtusely rounded, truncate, or slightly notched. This type is represented in the Oregon Jurassic by a considerable number of specimens, but not by so many as some other types. They vary somewhat in size, but are all small in comparison with the forms of the lepida and Huttoni type. This type tends to pass into the sibirica type more than into any other. PL XXX, Fig. 1, gives a nearly entire medium-sized leaf, showing a portion of its petiole. Two of the lo))es of this are shown slightly- enlarged in Fig. 2. Fig. 3 shows a part of the summit of one of the leaves having the most numerous segments. Fig. 4 represents a nearly entire leaf of the largest size. This is shown slightly enlarged in Fig. 5. Figs. 6 and 7 give a part of a leaf having the widest segments seen. This shows a portion of the petiole. The digitata type of Ginkgo is pretty wddely distributed in the Oregon Jurassic. It is found at localities Nos. 2, 7, 14, and 19. " This may have been simply an enorof Kollbiunner's in failing to observe that Heergave thisform specific rank. Bartholin in 1894 (Bot. Tidsk., Vol. XIX, pp. 9(), 97, 108, pi. iv [.xii], fig. 1 ), reduced it to a variety. — L. F. W. JURASSIC FLOHA OF DOrCJLAS COUNTY, OREG. 123 (iiNK(i(> Ih TToxi (Stciiil)cr^) Ilccr." V\. X.XX, Fi.us. s r.': I'i. XXXI. Fijrs. 1-:^. 1833. Ci/rlojitiris (lii/itiit Iliittoiil (Stcnil).) Ileer: Jiogcl's (iartcnlloia, Jaiug. XXIII, p. L'lil, ]il. ilccc\ii. lig. 4. FS76. Giiibjo Ihtifoi,; (SUtdI).) FTcit: FI. Fo.s.s. Arct., Vol. TV. Pt. T ^Bcilr. ■/.. Foss. Fl. SpitzlxM'geiis), p. 43, j)l. x, lig. 10. 1S78. Sdli.sliiiriii lliiftoni (Stcrnh.) Sap.: Plaiitcs .luaras.si(|U('s, ^'ol. IH, p. L".)'.), pi. clix [\.\.\i]. ligs. 1. .") ; pi. cl.x [xx.xii], fig. S. 1900. Giiikgii (Ugitiita roriiia lliitfinii (Stornl).) Sew.: .Iiir. Fl. Yorksli. Coast, p. 259, pi. ix, fig. L'. Tlie fonns which 1 g;r()up uiulcr tlic spet'itic luinie Huttoni are the most al)undant next to those classed as G. sihirica. At some locahties the Huttoni form is more abundant than any otlier. The leaf substance is thick and firm. The leaves have as a I'ule fovu' segments, which often show no trace of sul^division. Occasionally one or more of the segments may be divided by comparatively slight incisions, and more commonly they may show a notching of the ends. The segments are very wide in proportion to their length. They are widest above the middle of the segment and generally narrow somewhat near the tip, which is rounded off or may Ije truncate. When the segments are more tlian four, the plant approaches the digitata type. Most of the leaves are larger than those given by Heer.'' A very large leaf of this type that is found in a good many speci- mens and is the most common Ginkgo at locality Xo. 2 may be a new "Most autliuis include this form m6.(ligiia(a,-w\m-h Lindlcy and Hutton believed it to be, but Sternberg separated it, redeseribed it, and named it Cyclopteiis Hntlonl. Ileer, lunvever, retained it, as did also Saporta. Mr. Seward, in his -Jurassic Flora of the Yorkshire Coast, reduces it to a mere form, of which he finds and lijjures a specimen from .Scarboroufjh in the British Museum, No. V, 3.57.S. In the descripti(m of pi. ix, li;;. 2, he does not separate it from (1. (HgUatn, but on page 259 he treats it as a form. .M'ter the arrival of his work in America I called Profes.sor Fontaine's attention to this, and in his letter of Auf;ust 21, 1901, from which I have already quoted extracts, says: "The (iinkgos gave me a good deal of trouble. There seemed to be no way of dealing with them except as I did, or making them all G. d'ujitala. I think G. Iluttoni as good a species as can be made out of such leaves. It is a common form, and very few specimens occur grading toward G. diffilala." I tlicre- fore retain the species and confine tlie svnonvmv to those names that refer toLindlevand Mutton's plant. — L. F. W. ''FI.F0S.S. Arct., Vol. IV. I't. I,p. 10, pi. x.fig. 10: I't. II, pp. .-)9~l>(l. pi. v.lig. lb: pi. vii.lig. 1: pi. x, fig. ,S. 124 :\[F.S(^ZOT(^ FLORAS OF rNITED STATES. species. As. howexcr, there is no marked difference except size to separate it, I will describe it as a variety of G. Huttoni, witli the variety name tnagiiifolia. The normal Huttoni leaves are mnch smaller. PI. XXX, Mg. S. gives a normal leaf with segments of the most deeply cut kind. i'"ig. V) shows this slightly enlargetl. Figs. lU and 11 represent one of the smallei- leaves with broad segments. Fig. 12 gives one of the smaller leaves with proportionally longer segments, more elliptic in form. PI. XXXI, Fig. 1, shows a leaf with very wide, short segments. Fig. 2 gives a portion of a leaf which has a greater number of segments than is common and which approaches the difiitata ty])e. It is .shown slightly enlarged in Fig. 3. The plant is most abundant at locality Xo. 7 and is conunon also at locality Xo. 2. It occurs not rarely at locality Xo. 19. Ginkgo Htttoxi m.vgnifolia Fontaine n. var. PI. XXXI. Figs. 4-,S; PI. XXXII, Figs. 1, 2. The form distinguished as Ginkgo Huttoni magniJ'oUa is pretty con- stant in character and, as stated, may really be a new species. The length of none of the leaves is preserved entire. The width may be made out very well from some of the specimens and the length may be estimated. It differs from G. Huttoni in the great width and length of the segments. As in G. Huttoni, they are normally four in mimber. and in general form they agree well with those of the specific type. Some of the leaves, as made out from the specimens, must have been 12 cm. wide in their widest part, and probably were 9 cm. from base to summit. The petiole is strong and very long. A portion of one was seen 6 cm. long. Some of the segments of the leaf were obtained from 25-30 mm. wide. The nerves are strong and rather remote. PL XXXI, Fig. 4, shows the most complete leaf oljtainetl of the smallest kind. It does not have the ends of the segments preserved. It retains a portion of the petiole, which shows the stoutness. The petiole and one of the sides are shown slightly enlarged in Mg. 5. I'^igs. 6 and 7 show a fragment that may give an idea of the length attained by the segments, although their tips are wanting. Fig. 8 shows a fragment of a leaf from which an idea of the great width of the segments may be obtained. PI. XXXII, Figs. 1 and 2, give a fragment of a very large -IIKASSIC VUmX OF norOLAS COUNTY. OHE(;. 125 leaf ill which oiu^ side is ciitii-e, alTording an idea of the witlth of tlie leaf. The summit of tliis is not preserved. This plant is nnich the most common (!ink) h pida (Ilccr) I)n.; Tiaiis. Roy. Soc. CiiUiula, Sect. I\', \i>\. Ill, |>. N, |il. ii, li^. ■-*. The Icjiidit \y\>c of (iinkfjo, as aves, where they are rcducetl almost to petioles. The secondary segments are long and narrowly elliptical to linear in form, widest near the middle and tapering toward their lower and upper por- tions. This form sometimes tends to pass into the sibirica type by intermediate shapes that are rather common. The lepida type departs fui'ther than any other from that of G. dif/itnta. PI. XXXII, Fig. 3, gives a nearly entire leaf and one of the largest found. It shows the long narrow segments well. This is shown enlarged in Fig. 4. Fig. 5 shows a primary .segment with several ultimate ones, the Ijasal i)ai't not l)eing preserved. Fig. 6 denotes a primary segment with the ba.se not preserved, having shorter and proportionally broader ultimate .segments than is common. Fig. 7 shows this much enlarged. Fig. S gives a portion of a primary segment with three ultimate ones of the smallest size. Owing to the deep incisions of the leaf and its spreading habit, it is rare to find entire leaves of this type. This plant is most connnon at locality X"o. 19, where it is rathei' abundant. It is also common at locality X'o. 7. It occurs more rarely at locality Xo. 4. GiNK(;() snuHKA I leer. PI. XXXI 1 1. 1876. Ginkgo HiUrica Heer: FI. Fo.ss. Aret., \'()i. IV, I't. II (.lura-Fl. Ostsibiriens) pp. 01, 116, pi. vii, fie;. 6; pi. ix, fit;, ."ih: pi, xi (excl. \v^. 'M))\ \A. xx, fig.s. :\h. 6e; pi. xxii, fig. 3. 1885. SaJinhnria {(rinl-ffo) sihirira (Heer) Dn.: Truus, Koy. Soc. Canada, Sect. I\ , \»\. Ill, ]). s, ])!, ii. fijr. 1. 1L*<; MESOZOK" FLORAS OF UNITED STATES. Leaves. — Xuiiierous s| )ocii I loiis of (iinkgos occui- in the Oregon flora that soniotinies h-dvv llie character of the form called l)v Heer G. sibirica and sometimes of his (i. SclnnidlidiKi." tlie latter l)eing a smaller leaf with a smaller mmiber of segments; hut these graduate so often one into the other, by commonly occm-ring intermediate forms, that there is no satisfactory way of separating them. 1 will therefore discuss them as one species. As the sibirica form is the most common, and apparently the type, I will regard them all as variations of (r. sibirica. This Ginkgo is the most abundant and widely diffused type in the Oregon Jurassic, the leaves sometimes thickly covering the surface of the rock. If we neglect the intermediate shapes it is true we may select forms that are pretty widely separated, either as G. sibirica or as G. Schmi'dtiana. This is possibly what Heer did, not having so many specimens to illustrate the passage forms. In the description I will first note those l)elonging to G. sibirica proper and distinguish the smaller leaves as being of the Schmicltiana form. Possi):)ly the latter may be immature leaves of the former. The normal or true sibirica type of Ginkgo in the Oregon Jurassic has the following character: The texture of the leaf is decidedly thinner than that of any of the other types except that of G. Schmicltiana. The leaves are small, being sometimes about the same size as those of G. (ligitata, but generally smaller. The segments of ultimate order are numerous, slender, and formed by a deep incision of the leaf, but not so deep as in G. lepida. These segments are narrowly elliptical or linear in form and widest near their middles, but do not vary much in width throughout their length. PI. XXXIII, Fig. 1, gives a portion of a leaf with numerous narrow segments. Fig. 2 shows the basal part of a medium-sized leaf, with a portion of the petiole well preserved. Fig. 3 gives one of the largest sized leaves, but in a fragmental condition. This is shown much enlarged in Fig. 4. Fig. 5 depicts a portion of a medium-sized leaf, with a central segment, well preserved. This is also shown much enlarged in Fig. 6. " I'l. Foss. Arct., Vol. IV, Pt. II, p. 60, pi. vii, fig. 5; pi. xiii, figs. 1-2. I have not put Ginkgo Schmidtiana ill tlu! synonymy of G. sibirica, althmigli Pnifcssor Fontaine clearly indicates that tlic Oiogoti forms embrace Ixjtli of Ilecr's species and show all the necessary conncctiiif; links. As Heer de.scrihed G Sehnndtinna on p. 60 and G. sihirica on p. fil of his work, if they are consohdaled the resulting species would, by the rules, become G. Schmidtiana and not (!. .liliirica. This nia_v .vet have to be done, but rather than do it I prefer to regard the species as distinct, and tlie .\merican forms approaching G. SchiniJtiana as representiug the range of variation of tlic [lolyjiiorplious species G. sibirica. — L. K. W. JURASSIC FLORA OF DOFCiLAS COUNTY. ORKO. 127 Fig. 7 gives tlio trrniinal pai't of a leaf woW preserved. TIh^ Srlniiullinna type is, as stated, a smaller leaf, also with thin texture and with fewer ultimate segments, often with only four, othei'wise the chai'acler is much like that of the normal (/. sihin'fd. The SchniiilUiiiin forms are not so abundant as those of G. sibiriai. and they occur in the .same local- ities. Fig. S gives one of the SchmidlidiKi l(>aves having the narrowest idtimate segments, and showing a jjortion of the petiole. Fig. 9 gives one of the largest leaves of this type, with the widest segments. It is shown considerablN- enlarged in Fig. 10. l''ig. 1 1 .shows the basal poilioii of a leaf, with medimn-sized segments. , Seeds. — There are in the collections made at the Oregon localities a numl)er of small mit-like seeds that seem to be the seeds of some Ciinkgo. They are broadly ovate in form, and in size and shape agree exactly with the seeds attrilmted liy Heer" to (riiiLyn sihirica. These seeds vary but little in form and size. It is (juite probal>l(> that th(\v are in fact the seeds of Ginkgo sihirica, as the leaves of that species often occur with tliem. The leaves of other species occur with them also. They A-aiy slight 1\- in size and width, some being more narrowly elhp- tical than others. Fig. 12 gives one of the .smallest and most narrowly ellii)tical forms, shown enlarged in Fig. VS. Fig. 14 shows tw^o small elliptical seeds, also enlarged in Fig. 15. Fig. 16 denotes one of average size and shape, enlarged in Fig. 17. Fig. 18 gives one of the largest and the most brf)adly elliptical seeds ol)tained, enlarged in Fig. 19. These seeds are most abundant at locahty No. 7. They occur at localities Nos. 14, 18, and 19. GINKGOS OF ABERRANT FORMS. Ginkgo sp. Fontaine. PI. XXXIV, Figs. 1-12. Under this head I place certain Ginkgo leaves that do not plainly come under an\- of the previously described forms. They probably do not belong to any distinct species, l)ut are aberrant leaves of some of the species described in the preceding pages. I note them in ordei- «F1. Foss. Airt., Vul. IV, Pt. II, pp. (n-(V2, pi, xi, fi>- Seward on pi. ix, fig. 8, of his Jurassic Flora of the Yorkshire Coast, and named Ginkgo xvhitbiensis Nathorst. The segments are not so long and acute as those of Yorkshire, but they are not well preserved. It occurs at locality No. 7. Figs. 3 and 4 (enlargement) give a very small leaf with three elliptic segments, occuring at the same locality. Figs. 5 and 6 (enlarge- ment) give a larger leaf, hardly at all segmented, found also at locality Xo. 7. Figs. 7 and 8 (enlargement) denote a small leaf with four broad irregular segments. This also occurs at locality' Xo. 7. Figs. 9 and 10 (enlargement) represent another form found at locality Xo. 7. It is a petiole, bearing at its summit a pair of small leaves with three seg- ments each. Figs. 11 and 12 (enlargement) show a form from locality No. 18 that contains three narrow segments at the summit of a slender petiole. It resembles the form figured by Sew^ard, op. cit., pi. ix, fig. 4, and named Baiera Phillipsi Nath. Genus PHCENICOPSIS Heer. Phcenicopsis 'i sp. Fontaine. PI. XXXrV, Figs. 13, 14. Several fragments of strap-shaped leaves that at least suggest the presence of the genus Phoenicopsis were found in the Oregon Jurassic. They are quite rare, only two being found at locality No. 2, and one at locality No. 7. They are too poorly characterized to determine anything definite regarding them. They are strap-shaped fragments, which do not show either liase or summit. They do not narrow at all from one end to the other. Their margins are strictly parallel. The widest of them are 8-10 mm. wide, and the smallest are but a little narrower. In at least one case several fragments occurring together converge at one end, as if to a common point. The leaf texture seeflis to have been thin. There are so far as seen no well-defined nerves, but a fine striation is shown on the surface. These fragments suggest Heer's species Pkce- nicopsis speciosa." a n. Foss. Arct., Vol. IV, Pt. II, pp. 112-113, pi. xxix, figs. 1 (excl. Ic, Id.), 2; pi. xxx. JURASSIC FLOIIA OF DOUGLAS COrXTV. OUKC. 129 PI. XXXI\ , I'"i<2;. lo, sliows tlirce tVa^iiuMits which convortio and overlap at one end. Fi. 77, ])!. viii, fig. 1 . ls7."). Tii.i-'iUh laxus Phill.: Geology of Yorkshire, 'Ad ed.. p. 2.'51, lign. 64 on p. '2'\\ , pi. vii, fig. 24. 1110(1. Taxitcf! zamioidfs (Leek.) Sew.: Manehester Menioir.s, Vol. XLIV, Pt. Ill, No. S, p. 5; Jiir. FI. Yorksh. Coast, p. .•lOO, pi. .\, ilg. .5. Leckenbv described from the Oolite of Scarborough a small conifer- ous branch which he named Cycaditcs zamioidcs." Seward calls this plant Ta.vites zamioidcs. As the leaves narrow to the base like those of Taxus, Seward is no doubt right in regarding the plant as a Taxites. There are at some of the (3regon localities numerous imprints of leaves exactly like those of this plant. In most cases they are detached, and sometimes they thickly cover the surface of the rock. In one case they were found attached, as in Taxus. The leaves are small but have a thick leather- like texture. Tlie,y are about 4 cm. long and 1.5-2 mm. wide. They taper gradually to a subacute tip, and at base are abruptly- rounded off and attached by a ver>' short twisted petiole. The midnerve is very slender, but is sharply defined. PL XXXIV, Figs. 15 and 16 (enlarged) give a portion of a twig with several attached leaves. Fig. 17 shows a mmiber of detached leaves. PL XXXV, Fig. 1 depicts several nearly entire leaves. Some of these are shown enlarged in Fig. 2. The specimen represented in Fig. 3 pre- sents the appearance of a portion of rock covered with detached leaves. The leaves are most common at locality No. 19 and are found also at localities Nos. 1, 7, 17, and 18. "On the .sandstones mid slialos of the Oolites of Scaihoroiif^h, liy John Leckenby: tjiiarl. Jouni. Gcol. Soc. London. Vol. XX, p. 77, pi. viii, lig. 1. MON XI. VIII — 0.5 9 130 MESOZOIC IXOKAS OF UNITED STATES. Ucuus BRAGHYPIIYLLl-M Brungniait. Bkaciiyphyllum mamillare Brongiiiart. PI. XXXV, Figs. 4-8. 182S. Brachyplu/Iium momUlare Brongn.: Prodroiiie, pp. 109, 200." 1829. Thuites expansus ? Steriib. Phillips: Geology of Yorkshire, pp. 153, 167, 190, pi. X, fig. 11." 1835. Thuites expansus Sternb. Lindlej' & Huttoii: Foss. Fl. Gt. Brit., Vol. Ill, p. 49, pi. clxvii. 1836. Brachi/phyllum mammiUare Brongn. Lindley & Hutton; Oj). cit.. Vol. Ill, p. 99, pi. clxxxviii. 1837. BrachypliyUum mammUlare Bz-ongii. Lindley & Hitttoii: Op. <'it., "N'ol. Ill, p. 177, pi. ccxix. 1870. BrachypTiyUum Phillipsii Schinip.: Pal. Veg., Vol. II, p. 336.'' Several specimens of a plant that seems to jje identical with BrachypliyUum viamillare Brongn. were oljtained in the Oregon Jurassic. The plant is very rare and only small fragments of stems are nsiially found. These are poorly preserved and the leaves are generally distorted by pressure. The specimen given in PI. XXXV, Fig. 4 is a part of a branch about 12 mm. wide and 6 cm. long ^fat has the leaves of this character. A small portion of it is shown enlarged in Fig. 5. Fig. 6 shows a small fragment of a brancli with crowded leaves. This is shown enlarged in Fig. 7. Fig. 8 denotes a larger fragment that is of somewhat doubtful character. It is 9 cm. long and 2 cm. wide. It is an imprint made by a decorticated stem. It shows, irregularly placed and remote from one another, a number of obscure rhombic scars, and occasionally one that is elongate-elliptical, placed transverse to the axis of the stem. They can be seen distinctly only with the help of a lens. These scars " Brongniart never described or figured this species, but it is on page 109 of the Prodrome that he describes the genus BrachypliyUum, and places in it only this one species. This has proved sufficient to enable others to identify it, and is c()uivaU'nt to a description of the species, Brongniart, however, |)repan'd drawings of the plant, which were afterwards completed and puliHslicd by Saporta (Plantes Jurassicpies, Vol. Ill, p. 32S, pi. clxii [xxxiv], figs. 3-7).— L. V. W. '' The true Thuites expansu/! of Sternberg (Flora der Vorwelt, Vol. I, fasc. 3, p. 39, Tentamen, p. XXXVIII, pi. xxxviii, figs. 1,2) from the Stonesfield slate does not seem to occur in the Yorkshire Oolite, but both Phillips and Ijindlcy and llutton wrongly referred some of the Yorkshire forms to that species. — L. F. W. '■ Schiniper considered the form figured by Lindley and Hutton in tl>e Foss. Fl. Gt. Brit., Vol. Ill, pi. ccxix as a distinct species, and .Saporta's fresh drawings of Brongniart 's plant were regarded as establishing this fact. He states that both the figures of Lindley and Hutton arc of the same specimen, but Mr. Seward, who foimd the specimen in the Manchester Museum, docs not mention this.an conditions. This plant occurs at localities Xos. 12, 7, and 17. I'amily PIXACE.E. (ieiuis AliAlCAUrrFS I'lvsl. Araucaimtes ( s]). I^'onlainc (cone scale). PI. XXXV, Fin;. !). At locality Xo. 7 an a]>j)areii1 cone scal(» of Araucai'ites was found in a single specimen. It is quite convex, appearing hai'd and rigid, and was apparently quite thick. It is cuneate in form, expanding into a broad thick summit, the extreme tip being hidden. Toward the opposite end it narrows consideraljly, and at the end shows traces of former union with the axis of the cone. Its width at tlK> summit is 18 mm. and its length 25 mm. Genus PINUS Linnanis. PiNUs NordenskiOldi Hccr. PI. XXXV, Figs. 10-17. 1S76. J'inm NordensUlildl Ileer: Fl. Foss. Arct., Vol. IV, Pt. I (Beitr. z. Fo.ss. Fl. Spitzbergens) , p. 45, pi. ix, figs. 1, 2, 2h, :\, 'Ah, 4, 5, 5b, 6.' .At several of the Oregon localities there are a numV^er of long Pinus- like leaves that are always detached. They never appear in such numbers as the leaves of TaxitcH zamioides, but are moi-e scattered. They agi'ee well with the larger forms attributed by Heer to Pinus Nordenskioldi. Some of them, however, are a good deal larger than any given by Heer. These larger forms can not be separated from certain smaller ones found in the Oregon flora, which agree very well in size with the largest of Heer. These leaves are pretty siu-ely those of a Pinus and not of a Taxites. The}' narrow graduall)' to subacute tips and do not narrow to the base. In one case a fragment of the sheath may be seen attached to the base of a leaf, and two other leaves lie near this, as if they had once all belonged to the same bundle. There is a considerable variation in the size of the leaves. The largest leaves, none of which were seen entire, have a length of 115 mm. " I (io not inchuip tlic ('i)ch]>ifijs .\orilcnsl{iiil(li (Ileer ) Sclimalli. in llic synonymy of this species, heeause after carefully coinparinf; Schninlhansen's ligincs witli those of Ileer, I have grave doubt whether Sehnialhausen had this plant at all. His genus Cyclopitys seems to be <|uite distinct. — L. F. W. 132 .MESOZOIC FLORAS OF UNITED STATES. and a inaximuin widtli of o nun. Tlie niidnerve is strong and the texture of the leaves very thick and leather-like. They are often transversely wrinkled, no doubt from shrinkage. PI. WW. l"ig. 10 represents a gi"oup of leaves, one of them .showing traces of a sheath at its base. These are shown slightly enlarged in Fig. 11. Figs. 12 and 13 give portions of two medium-sized leaves. Figs. 14 and 15 represent the longest leaf seen, it not being entire. Figs. 16 and 17 denote the witlest fragment found. The leaves are most abundant at locality No. 19, and they are not rare at locality No. 7. Genus CYCLOPITYS Schmalliausen." Cyclopitys oregoxensis Fontaine ii. sp. PL XXXVI. Fig. 1. -2. A single specimen, and its reverse, of a whorl of Taxus-like leaves was obtained at locality Xo. 7. Schmalhausen has established the genus Cyclopitys* for the plants that have an affinity with Sciadopitys. The Oregon fossil resembles those that he calls Cyclojntys Nordenskioldi,'' but the leaves are broader and not so rigid, and are probably longer. The Oregon fossil shows only one whorl, composed of 18 leaves, some of the leaves being evidently missing from the whorl. The leaves are attached by thin narrow^ed bases. The stem is not seen. The leaves are rather thin in texture. Their precise mode of attachment can not be made out, as their bases are crowded and distorted. They seem to be attached by narrow^ed bases and not by petioles. Only the basal portions of the leaves are preserved. They seem to have been of about the size of those of Taxites zamioides. There is a midnerve in each leaf that is rather slender. This nerve sometimes appears as a slender cord-like nerve with a depressed line on each side, and sometimes as two closely placed nerves. This difference seems due to the fact that in the one case the lower surface of the leaf made the imprint and in the other it was made by the upper face. Schmalhausen'' regards Heer's Pinus Nordenskioldi as a Cyclopitys. This mav be the case with some of the forms that Heer has united in that a It seems probable that this genus belongs to the familj' Taxacese instcud of the family Pinacese. — L. F. W. ''Beitriige z. Jura-Fl. Russlands: Mem. Acad. Imp. Sci. de St. Peteisb., 7" si-r., Vul. XXVII, No. 4, pp. 39-41. «3 Op. cit., pi. xiv, figs. 6-8. f'Op. cit., p. 39. JTRASSIC FLORA OF DOFCiLAS COFNTV. OHFCF 1:^)3 species, foi- ho seems to have been rather lavish in his application of the name, Init it can hardly 1)0 assumed that the lonji Pimis-like leaves fjiven Iw Heer from some localities are a (yclopitys, or anylhiiiii like it, for they are much longer tlian any of the foi-ms that Schmalhausen gives of his CyclDinlij^ Xonlenskioldi. (Icinis SPHENOLEPIDIUM Floor. Sphenolepidium oreconense I'^oiitainc ii. sp. PI. XXXV], Fi- are apparently nearer to these forms than to any other previously described plants, they do not seem to be identical wiih any of the species of Samaropsis described by Heer, and their true position is doubtful. The two Oregon fossils differ somewhat in form, but evidently are the seeds of the same genus if not the same species of plant. They seem to have been winged seeds. The wings show only traces in a marginal ring, the fossil being mainly the seed or nucleus to which the wing was attached. Owing to the difference in their form 1 shall dis- tinguish them as form a and form b. Both of these seeds have a smooth bony surface. They are sur- rounded by a depressed line representing the attachment of the wing to the seed. Outside of this there is a narrow border, less than 1 nma. wide, which is all of the wing that remains, if it ever existed. Form a, represented in PL XXXVI, Fig. 9, enlarged in Fig. 10, is the larger of the two. It is ol)long in form, abruptly rounded off at one end, which seems to be the base. The same width is maintained to near the opposite end, where it terminates in a lancet-shaped tip. This seed is 5 mm. long and a little over 2 mm. wide. Form b, given in Fig. 11, enlarged in Fig. 12, is cylindrical in shape, slightly narrowed at the ends, and a Fl. Foss. Arct., Vol. IV, Pt. II, pp. 80-82. .UHASSIC FLORA OF DOrCU.AS COrXTY. OUK(i. 135 rouiidod off there. It is slightly eurve.l. It is f) inin. lon.ir and not (|uite 2 mill. wide. i Whatever their true position may be, they are evident ly very different from the other seeds found in the Oregon .lui'assie. MALE AMEXT OF CONIFER. I'l. XXXVl, Fi.u. \:>,. From U)cality Xo. 7 there was obtained a single specimen of a catkin- like object that seems to be the male ament of some conifer. It is poorly preserved and shows only an axis about 3 cm. long, with the scales of one side onl\- preserved. These are thickly crowded together and overlapping. They are thin in texture and ovate in form. PLANTS OF DOUBTFUL AFFINITY. Under this head I place a number of forms which are of more or less doubtful character. Genus YITCCITES Schimper aiui Mougeot." YucciTES HETTANGENSis Saporta? PL XXXVII, Figs. 1, 2. ISTO. Yitccifes hettangensis Sap. in Schiniper: Pal. Vi'g., Vol. II, p. 427. ISNG. Vuccites hettangensift Sap.: Plantes Jurassiciues, Vol. IV, p. 74, pi. ccxxxv [ix]; pi. ccliii [xxvii]. Several ribbon-shaped imprints were found, mostly at localities Nos. 2 and 7. They seem to be made by broad grass-Uke leaves, but may be caused by thin succulent stems. They have parallel sides and do not diminish in width in the portions preserved. The largest fragment obtained is 125 mm. long, with neither base nor tip preserved. One side or margin is imperfect, and it may have been wider than it now appears. Still, the width of 25 mm. is shown. No definite nerves appear, but irregular and rarely strong nerves are seemingly shown, " In view ot the general doubt on the part of leading paleobotanists as to whether the forms described under this generic name are really related to the living genus Yucca or represent Monocotyledons at all, I have not thought best to introduce the Angiospernis as occurring in our Jurassic flora. Personally I do not believe tliat they occur, and it is probable, as Schenk suggests, that if we ever ascertain tlie fnie nature of these forms we shall find them to fall into that general line of development that seems to lead from the Cordaitales ot the Paleozoic to the Ginkgoales. — L. F. W. 13(i MESOZOR' FLORAS OF UNITED STATES. most of the surface being finely striate. The apparent strong nerva- tion is prol)al)l\' a puckering from longitudinal folds in the leaf. The .shape and size resemble these features in Saporta's Yuccitcs hettangensis. PI. XXXVII, Fig. 1, gives the largest and best specimen ol)tained. Fig. 2 represents a small area of it enlarged to show the nerves. It occurs, as stated, at localities Xos. 2 and 7, but also at locality Xo. 19. rndetenninod leaf No. 1. PI. XXXVII, Figs. 3, 4. A few detached leaf-like impressions are found at localities Xos. 2 and 7 that are too poorly preserved to be definitely determined. They are found as detached scraps. Fig. 3 shows three of them, placed as if once attached to a common support, and Fig. 4 is a pen drawing of one of these. They are strap-shaped in form and narrow slightlj^ toward one end. The opposite ends of these fragments seem to be near their true bases, but the attachment is not visible. They seem to have been thick and succulent and to have had no definite nerves. There is an appearance of irregular nerves of unequal strength, but this is probably a puckering due to pressure. The general character of these objects is suggestive of the leaves, not bracts, of Willimnsonia gigas-{L. & H.) Carr., as given b_v Saporta." Saporta maintains that this Williamsonia did not have the foliage of Zamin gigas (L. & H.), as Carruthers assumed. Fig. 4 illustrates the best specimen found. Undeterinined leaf No. 2. PL XXXVII, Figs. .5, 6. One specimen of this leaf was found at locality No. 7. It is a long grass-like imprint, which has a vague fine striation and a seeming mid- rib. The latter is perhaps a longitudinal wrinkle. It has neither base nor tip, but is still 155 mm. long, with a width at one end of 15 mm. and at the other of 12 mm., indicating a narrowing. The seeming midi-ib shows no vascular ftssue and is vaguely defined. Possibly this is a narrow Yuccites leaf. It is shown in PI. XXXVII, Fig. 5, and a small area enlarged is shown in Fig. 6. « Plantcs Jurassiqucs, Vol. IV, pi. ccxlii [xvi], fig. 1. ' JURASSIC FLORA OF IK)1(;LAS COrNTY. OKKG. 137 Gcmis CARPOLITITrs Allioiii." CARi'oi.niirs oi.Ai.i.KNsis AA'ard n. sp/^ ]'i. XXXVII, Fi<,rs. 7, s. Two i)u1-lik(> ohjects were found al locality Xo. 7 that seem to be essentially the same, although varyin mil-liko objects wciv found, one each at localities Nos. 1 and 7. They seem to he the same species. They are. not unlike llie seeds of Giukiio hilolxi, but an- much hivs^er. They are bi'oadiy ellip- tical in form and are smooth on the surface and I'ounded at both ends, which do not diilVi' in foi-m. They ai-e ciuitc- convex, but the orip;inal surface seems to have l)eeu removed. The form given in PI. XXXVTl, ITij. 10. is 1 cm. long and '27 nun. wide in its widest part. That given in I'ig. 11 is ;17 mm. long and 27 mm. wide. CaKPOI.ITIU S ELOXGATUS FolUailU' 11. up. I'l. XXXVll. Fiu-. ]■_>. A single specimen of a narrowly elliptical nut-like oliject was obtained at locality No. 7. It is the same type of plant as C<>litlms orefpnensis, but is smaller and proportionally much narrower. The surface is smooth. It is slightly vmsymmetrieal, one longitudinal mar- gin l)eing more convex than the other. This may he due to distoi-tion. Its full length is not preserved, btit it is still 39 mm. long and 16 mm. wide. It was probably over 4 cm. in length, giving a great length in proportion to its width. Carpolitiius doxtglasensis Fonlaino n. sp. PI. XXXVH, Fiu;. 13. A single specimen of a peculiar pod-like organism was found at locality No. 19. It is not convex, but lies even with the surface of the rock and does not seem to have had nuich woody matter in its com- position. It has at one end, the probable base, what looks like a frag- nuMit of the stem to which it was attached, and at the opposite end an apparent double beak, as if it were a two-^'alved pod. The length is 2 cm. and the width in the widest part 1 cm. It is elliptical in form, narrowing gi'adually from base to apex. Whatever it mav- be it is certainly different from any of the previously described nut-like objects. I name it from Douglas County, in which it was found. 140 MKSOZOIC F1,()1;AS of rMTKl) STATES. Jur 1. roxcLVsioxs. The folhnvinji: i.s the coinplete list of ■assic : the ])lants found in the Oregon 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 26. :ii. 28. Marfhanlitcs crcctus (Beaii~) Sew.? Dick.soniM orogonen.sis P^mt. n. .-eds. 142 :\IESOZOIC FLORAS OF UNITED STATES. The following species and varieties have been found in the Oregon and OroA-ille strata and elsewhere in Jurassic beds: Sagenopteris Goepper- tiana, PferophyUum rajniahalense, Podozamitcs kinceolatus, Podozamites lanceolatus latifolius, Pinus Nordenskioldi, 5 in all. We find, then, that the Oroville and Oregon beds have in common 12 out of 37 fairly well-characterized species, nearly one-thii'd. This shows pretty conclusiveh' that whate^-er the age of these strata may he it is essentially the same. But these last-named 5 species are found in other localities besides the Oroville and Buck Mountain regions. Sagenopteris Gceppertiana, according to Zigno, occurs in the Lower Oolite of Italy. Ptcrophyllum rajmahahnse was first found in the Rajmahal series of India, which is held to be Liassic in age; l)ut it does not differ apparently from Heer's Pterophyllum Sensinovianum, so, if Heer's conclusions as to the age of the strata containing it are correct, we may assume that it persists into the Lower Oolite. Podozamites lanceolatus genuinus, or Podozamites lanceolatus simply, is, like Chido- pldebis xohitbiensis, probably a much abused type of leaf. It probaljly is not a species, but rather a type of leaf found in many species which lived in Jurassic times. The original is from the Lower Oolite and the form is probal;)l>^ more characteristic of that period than any other of the Jurassic. The broad form, var. latifolius, seems to be more common in the Oolite than in any other epoch. Pinus N ordenskidldi is apparently also especially characteristic of the Lower Oolite. This, too, is probably not a single species, but rather a type of Pinus leaf that was common in the Lower Oolite. It is quite probable that some of the forms placed in this species are really Taxites. There remain 25 species that, as yet, have not been found at the Oroville locality. These, taken in connection with some of the last mentioned as common to Oregon and Oroville, show a remarkably large proportion of plants common to the Oregon beds and the two widely separated regions, Yorkshire in England and eastern Siberia. The element common to Oregon and eastern Siberia might be explained by supposing that in Jurassic times land connection existed between Asia and northwestern America. It is difficidt to understand why so many forms should be common to England and northwestern America. In this connection it is interesting to note that several of the forms made ,ii'i;assic ki.oha of Doidi.As <<)rNiv. okfj;. 143 known by Iviclilhofcu from China exist in the ( )i-cl)i.s liait)iinu'ii.sis. :i. Thvrsopteris Murrayana. 4. Kufi'ordia Gopperti. .5. TiPiiiopteris major. 6. Tffiiiioptcris vittata. 7. Sagenoptori.s paucifolia. S. Ptilozaiiiitcs Li'ck(Mil)yi. !l. Xiisoiiia compta. 10. Podozamitcs lanccolatus. 11. Ctenis sulcicaulis. 12. Ginkgo digitata. ]'■]. Taxites zamioide.s. 14. Bracliyiihylliim mamillare. Fourteen out of the 37 well-characterized species not new. In common with the Siberian Jurassic there are the following: 1. 2. 3. 4. 5. 6. 7. 8. Cladoplilebis acutiloba. Tliyrsopteris Murrayana. Nilsonia orientalis. Nilsonia parvula. Ptcrophyllum rajmahalense. Podozamites pulchellus. Podozamites lanceolatus. Podozamites lanceolatus minor. y. Podozamites lanceolatus latil'olius. 10. Ginkgo digitata. 1 1 . Ginkgo Huttoni. 12. Ginkgo lepida. 1.3. Ginkgo sihirica. 14. Seeds of Ginkgo. 15. Pinus Nordenskicildi. 16. Bracliyphyllum mamillare. Sixteen out of the 37 determined species not new. The Oregon strata rival those of eastern Siberia in the development of Ginkgos, and it is a noteworthy fact that nearly all the more impor- tant species made by Heer from those beds have similar forms in the Oregon strata. This similar great expansion of Ginkgos is strong proof 144 IMESOZOIC FLORAS OF INITED STATES. that the Oregon ^strata are of the same age as those of Sibeiia. We may go further and take it as good proof that the l)eds are not older tliaii Lower Oohte. This large element of the Oregon flora common to both the Yorkshire and the Siberian strata may he taken as strong confirma- tion of Heer's belief in the identity of the age of the two formations. There are in the Oregon flora, l)esides PterophyUum rajmahalence, two previously known plants that point to a somewhat older age than Lower Oolite. They are Ptcrojiln/lluin aciudic and Xil.sonia /ifeni/iln/l- loides. Both of these are given ))y Xathorst as found in the Rhetic of Scandinavia. The latter has not hitherto l)een found in strata younger than the Rhetic. Pterophlyllum (vqnalc has been noted In* Schenk as found in the beds of the Tumulu coal field of China. " These strata are, he thinks, of Lower Oolite age.'^' There can Ije no doubt, in the opinion of the present writer, that the Yorkshu-e Lower Oolites, the strata of eastern Siberia and of the Amoor, made known by Heer, and the Oregon beds are of the same age. The only question is, What is that age? The investigations of the English geologists would seem to have settled the question for the Yorkshire formation. ZeiUer, in his paper discussing the age of the fossil flora of the Altai made known by Schmalhausen, in a footnote ta page 478, ' states that it is questionable whether tlie Siberian and Amoor plants described by Heer are really Lower Oolite in age. He thinks that the resemblance of this flora, in a number of its elements, to that of the Rhetic of Scandinavia makes the question an open one and that the age may be Lower Lias or even Rhetic. He takes pains, however, to state that he does not maintain that Heer's conclusions are erroneous, but that the question of age in the case of these strata merits further stud}'. The key to the whole matter is the correctness of the determination of the age of the Yorkshire beds. So far as my knowledge goes no one has questioned the correctness of the conclu- sions of the English geologists regarding the age of the Yorkshire strata. That being established as Lower Oolite would certainly indicate a simi- " Schenk, Jurassic Plants of China, pp. 247-248, pi xlviii, fig. 7. fcOp. cit., p. 26.5. <• Remarque.s sur la (lore fossilp de I'Altai a piopos ties dcrnirres drcouveites pult'ohotaniques de MM. Bodenbender et Kurtz dans la R^publique Argentine, par. M. R. iieiller: Bull. Soc. Cieol. de France, 3" s^r.. Vol. XXI\', Paris, 1896, pp. 466-487. COLLKCriONS FROM VAKIOIS LOCALlTiKS. 14") lar aiio foi' the Siberian beds and also for those of Oregon. The sinii- laiily of a nnniber of the forms to those of the Lower Lias and the Rlietic may be exjihiined l)y a survival of a portion of the Rhetic flora into the Lias and the Lower Oolite. It nuist !)e rcMnembemi that climalal ronditions were apparently remai'kably unifoi'tn 1hr()U,chout the .lurassic, a condition very favorable foi' the persistence of types. The I'esem- l)lance between the plants of the Lower Lias and Rhetic formations and those of Yorkshire has lonji l)een known. OTIIKK PL,AXT-1$I:ARI>(; beds IX TIIF. JUHASSIC, oil FORMTX(; TITP, TUANSITIOX TO TIIK LOWKT. CUICTAC'KOUS. On August 8, 1884, Dr. A. C. Peale, while making investigations in northern Montana, collected -a small specimen bearing the distinct impression of a coniferous leafy twig. It was foimd on the east slope of the Bridger Range, north of Bridger Creek, 4 miles northeast of Bozeman, in about latitude 44° 44' N., longitude 110° 44' W., in strata regarded by him as Jurassic, and the specimen was so labeled. It is in a limestone underlying the green sandstone described in Hayden's Annual Report for 1872, on page 475 ("Last foot of Bridger Peak, etc."). The plant proves to be the Sequoia Reichenbachi, which certainly extends to the base of the Cretaceous and has its ancestral forms in the Jurasi(-. In the summer of 1884 Mr. Henry D. Woolfe sent to the Smith- sonian Institution from Cape Lisburne, Alaska, two boxes of coal, some rock specimens, and some shales bearing leaf impressions. These last were referred to the Department of Fossil Plants of the National Museum, where they remained some time, but finally, on September 30, 1885, they were sent, along with a number of other undetermined collections, to Prof. Leo Lesquereux for determination. In due time Professor Lesquereux submitted his report, or rather a series of reports, which were compiled and edited by F. H. Knowlton and published in the Proceedings of the United States National Museum, Vol. X, pp. 21-46; Vol. XI, pp. 11-38, pi. iv-xvi. The plants from Cape Lisburne are described in Vol. X, p. 36, and Vol. XI, pp. 31-33, and figured mostly on pi. xvi of Vol. XL A larger and better collection than the one made by Mr. Woolfe and from the same general rcigon has recenth^ arrived. From the meager data that have been furnished it is learned that it was made M»N XLVIII — 0.5 10 14(5 MESOZOIC FLORAS OF INrrKD STATES. by Mr. H. D. Dumars 80 niilos east of Cape Lislnu'iie in 1890 and was donated to the National Museum by Mr. A. G. Maddren, of Seattle, \\'a.sh. It was taken from the Convin coal mine. The rock is similar to that of the Woolfe collection and some of the species are the same, but there are others and very different ones. The specimens have a decidedly Lower Cretaceous, or even Jurassic, facies. On December 15, 1900, Dr. T. W. Stanton turned over to me a specimen containing a well-marked impression, with its counterpart, of a leaf which was collected by Mr. A. C. Spencer in August of that year on Nikolai Creek near Nikolai, in the Copper River region of Alaska, in strata supposed to l)e Cretaceous or Jm-assic. In February, 1901, Doctor Stanton placed in my hands a small collection of fossil plants collected by Mr. Diller's party the previous season in Curry County, Oreg., at a locality in the Port Orford cjuad- rangle, and labeled by Mr. Diller Jurassic or Lower Cretaceous. In March of the same year Doctor Stanton referred to me a speci- men collected by Mr. Ernest G. Locke, of Seattle, on Herendeen Bay, Alaska, labeled as coming from the "coal measures" of that region. The specimen showed the im^^ression of a cycadaceous leaf. Another collection from Alaska made in 1901 by Mr. F. C. Schrader was sent me by Doctor Stanton on November 25 of that year. The following is the list with field numbers attached. The trunk of a tree numbered A at the end of the list was not sent to Professor Fontaine. The label states that it was "collected by Mr. and Mrs. J. H. Holland in 1901 in Iliamna oil region, Iliamna Bay, Cook Inlet, Alaska; appar- ently Mesozoic." It is probably coniferous, and has a warty exterior as if the bark was partially preserved." List of localities of fossil plants collected by F. C. Schrader durinej the season of lOUl along the one hundred and, fifty-second meridian north of Arctic Circle and on the Arctic coeist of Alaslca. (Horizon, probably Mesozoic.) .544. Fos.sil plant stems in dark, dirty ;her terms of the cliussificiition. Where they have already occurred in the larjier report reference is made to the descriptions and .synonymy. 148 MEt^OZOlC FLORAS OF UNITED STATES. REPORT ON THE VARIOUS COLLECTIONS DESCRIBED ABOVE. By AVm. M. Fontaine. 7. PLANTS Fli(JM crEIiY COUXTY, OliEG. Mr. J. S. Diller collected 17 specimens of fossil plants from the Forks of Elk River, in the Port Orford quadrangle. Curry County, Oreg. The horizon is given as Jurassic or Lower Cretaceous. The plant mate- rial is very poorly preserved. Most of it is not determinable. All of it is in the form of small fragments, which have apparently been floated some distance from the place of growth. Hence the determination of all the fragments can not be positive. The following specimens are the only ones that possess character deserving notice. DiCKsoNiA oREGONENsis Fontaiiic?." PI. XXXVin, Fig.s. 1, 2. One of the rock specimens from Curry County contains a small fragment of a fern that resembles Dicksonia oregonensis Font., of the Buck Mountain Jurassic beds. The specimen is a small fragment of the end of an ultimate pinna. This fragment contains several fructified pinnules. The sori are large and globose in form. They strongly suggest the identity of the plant with D. oregonensis. The latter and Cladophlebis vaccensis are not uncommon in the Jurassic strata occurring in the vicinity of Buck Mountain in Douglas County, Oreg., and have lieen described in this paper. Owing to the small amount of material from the Curry County beds the identification can not be positively made. The specimen is shown natural size in PL XXXVIII, Fig. 1, and enlarged in Fig. 2. Thyrsopteris Murrayana (Brongiiiiirt) Heer?* PI. XXXVIII, Figs. :i, 4. On one of the specimens there occurs a small bit of a fern that resembles Thyrsopteris Murrayana. It is a part of the termination of an ultimate pinna, and carries several pretty well preserved pinnules. The form, texture, and iiei'vatioii of these pinnules indicate strongly the presence of T. Murrayana. Hut this part of a fvvu fi-ond is not well « For the description of this species see pp. 55-56. — L. F. \V . '' For the synonymy of this species see pp. 61-62. — L. F. W. PLANTS FROM CrilKV COINTV, OKWi. 145) iidaptod for detorminatioii, and llic aiiioiml of material i;-! too small. This foni occurs (luito commonly in the Buck Mountain Jvu'assic. PL XXXV'lll, l''ig. 'A, sliows tlK> specimen natural size, and l'i nervation is fine, close, and approximately parallel. This is another of the plants that are probabl.y common in the Curry County and Buck Mountain floras, for P. lanceolatus minor is found in the Buck Mountain localities. The material from Curry County is not sufficient to permit a positive identification. PI. XXXVIII, Fig. 11, shows the most complete leaflet natural size, and Fig. 12 the same enlarged. Otozamites oregonensis Fontaine n. sp. PL XXXVIII, Figs. 13, 14. Two fragments occur in this collection, one on each of two I'ock specimens, that seem to be a new species of Otozamites. The fragments do not suffice for a complete diagnosis, but perhaps justify a description of the plant as a new species. The most complete and largest specimen is depicted in PI. XXXVIII, Fig. 13, and a smaller and less complete one is given in Fig. 14. The « For syiioiiyiiiy see [>. 111. — L. F. W. I'l^AMs FROM ( nan' corNTV. okkg. 151 larger loaf is lotii in liaif l()ii base and fork I'cpeatedly, so as to fill the lamina of the leaf. In form and size th(> leaves resemble those of Oldzninilcs Kli/).' W'ealden of Enjiland. They most resemble tho.se of the form figured on pi. vii, fig. 0, of Seward's Wealden Flora, Pt. II, but the leaves now in question narrow more toward the Ijase. Taxites zAMiouiES (L('(k(nil)y ) Seward." PI. XXXVm, FIrs. 15- is. The principal specimen regarded as representing Taxites zamioides w a small bit of a twig with the bases of several leaves attached to it. Enough of the leaf is preserved to show a very close agreement with the T. zamioides of the Lower Oolite of Yorkshire, England. On some of the specimens several detached leaves occur that are nearly entire, and which agree well with those of this conifer. This conifer occurs at some of the .lui'assic localities in the vicinity of Buck Mountain, Oregon, not rarely, and has been described by the writer in this paper. This is shown natural size in PI. XXXVIII, Fig. 15, and enlarged in Fig. 16. A single nearly c(Hiiplete leaf found detached is represented in Fig. 17, and Fig. 18 show's this enlarged. The above-described plants are all that can l)e made out with any degi'ee of certainty in the collection from Curry County. They indicate with a high degree of probability that the strata which yield them are of the same age as the Jurassic strata of Douglas County in the vicinity of Buck Mountain. The abimdant plant fossils of these beds show that they are of Lower Oolite age. « For the synonymy- of this sppcics spc p. 129. i:)-J MJiSOZOIC FLORAS OK I'MTKI) STATKS. 3. PLAXrs FROM lIEREyDEEX BAY. ALASKA. Pteeopiiyllum alaskense Fontaine n. sp. PI. XXXVIII. Figs. I'.t, 20. This is a fragment of a cvead leaf which is most prol)ably a Ptero- phylhim. It belongs to that type of Pterophyllum of which P. rajmaha- Iciise jNIorris is the most important exampl(\ This, however, seems to be a new species. The leaflets are remote and attached to the side of the axis or midrib by the whole of a somewhat widened base. This is tlie most important difference distinguishing this plant from P. rajn/dlKih rise. The leaflets are oblong in form, with very ol)tiise ends. As is common in Pterophylla of this type, they show some variation in width even in the same compound leaf. The length for the average size is about 14 mm. and the width above the ]:)ase 4 nmi. The nerves are not well shown, but seem to Ije slender and lumierous, parallel, and about 12 in number. They may be more numerous in the wider leaflets. The texture of the leaflets is rather thin. Owing to the expansion of the bases of the leaflets they touch one another in those parts and are even united there with a U-shaped sinus. The largest leaflets may be twice as wide as the narrow- est. As in the case of P. rajmahalense, the variation in size is mainh' in the width, the length remaining pretty constant. The specimen is shown natural size in PI. XXXVIII, Fig. 19, and a j)()rtion of it enlarged in Fig. 20. Only a single specimen of this plant was found. It was collected by Mr. Ernest 8. Locke from the ' ' coal measures ' ' of Herendeen Bay, Alaska. Doctor Stanton states that the invertebrates accompanying this plant are Aucella crassicollis, a Lower Cretaceous species. 3. PLANTS FROM THE COPPER RIVER REGION, ALASKA. Sagenopteris alaskensis Fontaine n. sp. PI. XXXVIII, Fi- 21. This plant is found in onl>- one specimen, which is a nearly complete leaf, apparently a lateral one of the cluster characteristic of Sagenopteris. It was collected by Mr. A. C. Spencer from the Copper River region of Alaska, on Xikcjlai Creek near Nikolai. It is apparently one of the lateral PLANTS FROM ALASKA. 153 leavos of the fj;roup, as it is unsymmptrical in sliajjc. Tho leaf is broadly elliptical in form, narrowinfj; to an obtuse lip. The basal pai't of tlie leaf, on the left-hand side, is not entire, but the margin on this side was evi- dently not so sti'oiiiily curN'ed as the rit leaves a v(>iy distinct impression, aJthouii'h it is preserved in a coarse ijrit . The most complete leafl(>t has a leiii!;th of 7 cm. It is widest near the base, whei'e it is 4 cm. wide. The midnei-ve is fiat and obscure and it does not exist for more than one-third of the length of the leaf. The secondary nervation can not be made out. Tliis plant is c{uite near Sagctioplcris (lopprr/idiid fi-om the Lower Oolite of Italy," which is common also in the Jurassic formation of the Buck Mountain re'ielding the two collections previously made. All of these collections have been turned over to me for examination and the present paper gives the results. DESCRIPTIONS OF THE SPECIES. Order RHODYMENIALES. Family RHODOMELACE.E. Genus CHONDRITES Steinberg. Chondiutes filiciformis Lesquereux. 1888. Chondrites filiciformis Lx. : Proc. U. 8. Nat. Miis., Vol. XI, p. 32, pi. xvi, fia;. 1." "Professor Fontaine considers this specimen too vague for determination. It is, however, quite clear on the stone and Lesquereu.x's figure represents it fairly well. It may stand as a problematical organism. — L. F. W. 155 PLANTS FROM ALASKA. FMlm.. PTERIDOPHYTA. (Ferns au.! Fen. Allies). Order FILICALES (Ferns). Family CYATHEAC'E.E. Genus DICKSONLV L'll.'riti«>i-. DicKsoxiA Sai-oktana 11(mt. PI XXXIX, Fiij:.s. \,-2. Ostsibiricns u.ul .1. A.nurhindcs), p. M". I'L n^'^' "-- '• ' ' ' fig^- '•'*'•-' ^- . Ti Tp.n,i..reux- Proe U.S.Nat.Mus.,Vol.XI, p -V^ Cat U. S. Nat. Mus., No. 243.5. On -v suvtll fragment of rock collected by Mr. Woolfe there ;s an • n^ a ^e n tlv^ agrees well with Dicksonia Saportana as deternnned ;:Ct1t:m ;;: ^uiSsic of BiUena. Indeed, ;n ^- .^^ ^^^^ ^ ; n.uiles and in the rather uncommon nervation this fossil i. exacth 1 ke p nt;it 1 t ^ma with a number of ultimate ones, in which the termina- tfoi a e wanting. Most of the pinnules are more or less lacerated, and :^;:^:; ca^ Les,uereux to make an ----^^•— ;;,^ the plant. He regarded it as Aspler^m ^^^^^-^T^^ ^^ ^lyn.e pinnules, however, have no resemblance to this plant. Onl. specimen^^^^uncl^ 1, gives the fragment, and Fig. 2 a pinnule mag- nified. Family POLYPODIACE^. Genus ONYCHIOPSIS Yokoyama. Onvcuiopsis psilotoides (Stokes & Webb) Ward n. con^b. PL XXXIX, Figs. 3-6. 18^4 Hymenovteris psitotoides Stokes & Webb: Trans. Geol. Soc. London, 2d ser.,. Sussex [revised edition], p. 5.5. 1836. CheilantUtes Mantemi (Brongn.) GopP-^ ^Y^t- ^'^- ^^'«^" ' p. 23L 1 :')(5 MESOZOIC FLORAS OF UNITED STATES. lS3y. CheiJaiifhitcs ileiitlviilatus V. A. Jvocnier [non (Broiigii.) Gopp.]: Vprst. d. Norddeiitsch. Oolithen-Gebirges, Xaclitrag, p. 9, pi. xvii, fi<^. la. 1S43. Confervites fssus Dunk.: P]\)iriaiii d. liolicrcn Gowcrlyscliulf in Ca.s.si'l, 1843- 1S44, p. 5. 1846. Conferritfs ibfius Dunk.: Moiiugr. d. Xorddculsch. Wealdenbilduiig, p. 1, j)l. i, lig. 1. 1846. Sphcnoptens Romeri Dunk.: Op. fit., p. 3, pi. i, figs. 3, 4, 4a, 5. 1846. Sphenopteris tfjura Dunk.: Op. oil., \). 3, pi. viii, fig. 5. 1865. Microlepia ManteJli (Bronga.) Etl. : Farnkniuter dor Jctztwelt, p. 216. 1SG7. Sphenopteris antipoduin Tate: Quart, .lourii. Gi-ol. Soc. London, Vol. XXIII, p. 146, pi. vi, fig. 3. 1S8S. Aspkliuni 0(rsff'di Heor. Lesciuercux : Proc. U. S. Nat. Mu.s., Vol. XI, p. 32 in part, <|uoad Cat. U. S. Nat. Mus., No. 2434, Lesquereux's Nos. 913-915." 1894. Onychiopsis Mantelli (Brongn.) Sew.: AVcalden Floia, Pt. I, p. 41, figs. 4, 5 on p. 50, fig. 6 on j). 52, pi. ii, fig. 1; pi. iii, figs. 1-4. Four specimens of ;i fern with very slender incisions are found in the collections of Woolfe and Duniars. Three of them are in Mr. Woolfe's collection. These were identified by Professor Lesquereux with Aspidinm Oerstedi Heer. Lesquereux must have given this collection a ^•ery hasty examination, for this plant has not the slightest resemljlance to .1. Oerstedi. The three specimens collected by Mr. Woolfe appear 1o belong to the upper part of the compound pinna or frond. One of tliem is given in PL XXXIX, Fig. 5. Fig. 6 represents a magnified ultimate pinna of the same in which the pinnules are reduced to loljes. The fossil collected by Mr. Dumars apparently belongs to a lower portion than these. This is given in PI. XXXIX, Fig. 3, and Fig. 4 represents, magnified, a portion in which the lobes have l)ecome pinnules. This form is a fragment of a compound pinna, or of the frond, which toward its termination passes into the form given in Fig. 5. These specimens show that the plant has the following character: The ultimate pinnjc make an acute angle with tlie penultimate rachis, which is apparently winged l)y a decurrence of the lowest pinnules. They are narrowly oblong and have linear, almost threadlike, acute pinnules, which are set on very obliquely and united at l)ase, so as to make the rachis of the ultimate pinna winged. The.y diminish in size « Unfortuniitely nil tlic specimens referred by Professor Ijesquereux to the same speeies were given the .same number in the eatalofjiie of the United States National .Museum, and as Professor Fontaine refers the speci- mens to different speeies tliere would l)e no way of indi<'atinf; his determinations had not Les(|uereux altaelied to each imprint a private number of his own. I am therefore obliged to (juote his private numl>ers in addition to the Mu.seum number. — L. F. \V. PLANTS IKOM ALASKA. 157 toward the ends of the ultiinalc piniue and -aw more and more united, so that they pass to lobes and finally to teeth. The ol)li(|uity of insei'lion of the pinnules makes the bases of the nltinialc |)iniKr wed^c^ sliapiMl. Towai'd the termination of the comixjund |)imia, or fi'ond, the ultimat(> pimue become redurecl to pimiules. In such parts the peiniltimate pinna% now reduced to ultimate ones, are nuicli elon,u;at(Ml and toward their ends have the pinmdes passins Mnd teeth. The nervation can not be made out, as all the specimens ai'e preserved in a sandstone. No fructification is sliown, and the plant nnist be determined from the character of its sterile parts. It tliffers somewhat from the character of Onijchiopsis psilotoidcs as given in most of the specimens hitherto descriljed, l)ut the vai'iation does not api)eai- gi'eater than the limits of the species. The variation is found in the facts that the Alaskan fossil has the pinnules, lobes, and teeth more closely placed than in most of the forms of 0. psilotoidcs and that the lamina^ of the foliage is in greater proportion to the nervation. The crowding may be due to a creep of the rock and to pressure, for these features appear in the specimens. This plant resembles (). clongata (Geyl.) Yok., a Jurassic fossil, but is more delicately incised. It probably lies between it and 0. psilotoidcs. It is also near a number of fossils described in Monogr. XT. S. (Jeol. Surv., Vol. XV (The Potomac or Younger Mesozoic flora). It especially reseml)les Thyrsopteris angustifolia Font., the form given in Fig. 8 being much like those depicted in that work on pi. xlv, fig. 3, and pi. xlviii, fig. 2. But the pinnules and lobes of the Alaskan plant are more slender than even these. Genus CLADOPHLEBLS Brongniart. Cladophlebis vaccensis Ward." PI. XXXIX, Figs. 7, N. 1888. "i Pecopteris denticulata Ileer. Le.s(ni(M('ux: Proc. V. S. Nat. Mus., Vol. XI, p. 32. On one of tlie rock fragments collected by Mr. Woolfe there is an imprint of fern which has the Museum \o. 2526 and Professor Lesque- " See pp. 6f)-()S. 158 MESOZOIC FLORAS OK IMIKD STATKS. reux's Xo. 9091). This is given by Lesquercux as Pccoptcris dcnticuUita Heer. He gi\-es no description and no figure of it. The imprint shows several imperfect ultimate pinn:r cariying a nunihci' of mostly mutilaled pinnules. The j)iiuue are detached, hut so placed as to show tlial they were once attached to a common rachis. Enough, however, of the char- acter of this plant is shown to make it most probal)le that it is identical with CIndophlcbis i^acccnsis, found in the Jiu'assic (Lower Oolite) flora of Douglas County, Oreg., and descrilied on page 66. Only one specimen of this fossil occurs in the collections. It is represented in PI. XXXIX, Fig. 7, and one of the piimules with its attachment to the rachis is shown in Fig. 8. Cladophlebis alata Fontaine. PI. XXXIX, Figs. 9-11; I'l. XL. 1SS8. Aspidium Oerstedi Heer. Les(|uereux: Proc. L. S. Nat. Mus., Vol. XI, p. 32 in part, quoad Cat. U. S. Nat. Mus. No. -24.34, Lesqiiereiix'.s Nos. 910b, 910c, 911b, 912, 916, 917. 1«88.? Pinus staratschini Heer. Lesquereux: Proc. U. S. Nat. Mus., Vol. XI, p. 32. 1889. CladopJihhis alata Font. : Potomac Flora (Monogr. V. S. Geol. Sin-v., Vol. XV) , p. 77, pi. xix, ligs. 5, 5a. 1889. Pecopteris strictinervis Font.: Op. cit., p. 84, pi. xiii, figs. (5, 6a, 7, 7a, 8, Sa; pi. xix, figs. 9, 9a; pi. xx, figs. 3, 3a; pi. xxii, tigs. 13, 13a: j)i. elxx, figs. 5, •5a, 6, 6a. The most common, and perhaps the most characteristic fern of the two collections, is one of those that Lesquereux identified with Aspidium Oerstedi Heer, although it is entirely different from that plant and the others of Woolfe's collection that he placed in that species. Some of the larger rock fragments contain a numl)er of imprints. The amount of material enables one to get a pretty good idea of the character of the fossil. The specimens seem all to l)elong to parts pretty high up on the pinna\ The most complete specimens show a considerable portion of an antepenultimate pinna, which carries portions of several penul- timate and ultimate pinnje containing a number of pinnules. WluMlicr or not this represents the frond or only a compound pinna cnii nol be determined. It is probably only a pinna. It shows that the frond must have been of consider'able dimensions and that the plant was probably arborescent. The rachises are strong and rigid. The primary and secondarj' rachises of this specimen, given in PI. XL, Fig. 1, seem PLANTS FKOM ALASKA. If)!* to have hoon a raised conl-like line, runiiiiip; down tlie centei- of tli(>ir uppei' sui'faco. The ])i'iniai'v i)inn;r of this speciincii go off fi'oiii lli(> main I'achis op])osile to one another. I'h .\L. I'i end of the comijound pinn;i t h;in the part I'epresented in Vig.. I. 1( is also a poi't ion of an ante|)(MUihimate ])inna. In this the ultimate pimue aiv nuich I'educed in size, and the .same is ti'ue of the pinnules; 'rii(> latter are more united and lend to pass to lobes and teeth. Many of tlu^ pimiules and lobes in this foi'in are much nuitihiled and do not show tlieii' time sha|)e. 1*1. .\L, ]''if>;. 4, shows the lower part of a piniuile enliiixcd. 'The foi'm fiiven in PI. XL, Fig. 5, probably i'e{)resents a stage between these two. ThLs gives one of the ]ien\iltimate pimue more complete. PI. XL, Fig. (), shows a jiinnule enlarged. PI. XXXIX, Fig. 9, probably represents a ])oi-tion of a penul- timate pinmi from lower down on the compound piiuia than any of those shown in PI. XL, Fig. 1. Tn this the pinnules aic inoi'e separated than in any of the other spe<'imens. They ai'e also larger, and the largest of them show serrate teeth. These can be seen distinctly only with the help of a lens. They ai'e shown in the magnified pinnule, PI. XXXIX, Figs. 10 antl 11. It is probable that still lower down the teeth become more pronounced and take the character shown in ('lodophlebis oJota. This plant seems to l)e identical with two fei'iis first found in (he Potomac, or Lower Ci'etaceous foi'mation. They are (IdduphUhi^ (ilata and Pecoptcris stricti7tcrvis. These specimens make it most probable that the two ferns from the lower Potomac of Virginia, called by the present writer Cladopklebis alata and Pecoptcris sirictincrvis. nre the same. Pecopteris stricfinenHs represents upper and terminal portions of the fi'ond and compound pimue. ('hiilojililvliis (ihila is the foi'ui found lower down. The larger pinmiles of the fossil represented in PI. XXXIX, Fig. 9, are forms establishing a passage from the nlata to the stnclinervis type. Most of the specimens belong to the stnclinervis type, and none with dentation so pronounced as that in ('.dhila were found. Some of the pinnules of Pecopteris sirictinervis, as seen in the \'irginia Potomac, show a toothing similar to that found in the foi'ui depicted in l"ig. .'1 The following description of the fo.ssil nuiy be given: The plant was prol)ably arl)orescent with strong rachi.ses aiul wide spread of foliage. Tlu^ epidermis seems to have been firm .and dui-able, 160 MESOZOIC FLORAS OF rNITKI) STAl MS. so that tlie jilant niattei' of the piiiiiulos is usually well ])rescrvod and leaves a sliiuiuir film on Ihe stone. The i-achis of the ultimate pinna is winged 1)V the decui'i'ence of a pimuile, oi- lobe, placcMl in the angle l)etween the lower side of the base of the ultimate rachis and the penultimate one. The larger pinnules in PI. XXXIX, Fig. 9, the dentate ones, may he regarded as normal for the frond. As stated above, they diminish in ascending on the frond, and also toward the ends of the ultimate jjimue, becoming entire and nioi-e united, until they pass into lobes and hnally into teeth. The general character of the noi'mal pinnules is maintained until they are reduced to lobes and teeth, when they become proportion- ally broader at base, taking more or less of an ovate form. The normal pinnules are narrowly oblong and acute. They are decurrent at base and united with the next lower ones, forming a narrow^ wing on the idtimate rachis. On the upper side of the base they are slightly constricted, the constriction being more pronounced the deeper the toothing. The larger and lower pinnules have minute teeth that are acute and inclined strongly toward the tips of the pinnules. They often have a spiny look and can not be seen distinctly without the help of a lens. The pinnules are generally straight, but may l)e slightly falcate. They go off obliquely from the ultimate rachis and are inclined forward toward its end. The midnerve of the pinnules is rather slender and continues to near the end of the pinnules. The lateral nerves go off pinnately and very obliquel}'. They are straight. In the toothed pinnules the lower ones are once forked, the rest are single. In the entire pinnules and lobes they are single. This plant has a good deal of resemblance to Aspidium montanense Font." of the Kootanie strata of Great Falls, Mont. But the Montana plant has the pinnules, lol)es, and teeth more obtuse and not so strongly inclined forward. PI. XL, Fig. 7 probably repi-esents a portion of a penultimate pinna from the upper part of the compound pinna, where the ultimate pinnse carry pinnules reduced to lobes. It shows the slender elongate form of these. Figs. 8 and 9 show enlarged pinnules of this. The fossil Lesquereux determines as Pinufi! Staraischini does not seem to be Pinus. It looks like the rachis of Ckidoplilebis alata. ' Doscriptioii iif some fossil plants froiii tlie Groat Falls <-i>iil liclil of Mdiilarui : I'idc. U. S. Nat . Mils., \'ol. XV, p. Am, pi. I.\,\.\ii; pi. l.wxiii, nj;s. 1, la, 2, 3. 3a. PLANTS FROM ALASKA. 161 Clauophlebis Ilt'TTONi (Duiikcr) Fontaiiu' ii. coinl). PI. XLl XLIII. 1846. \t'iir()j>tfns Hiiftoni Dunk.: Monou:r. li. Xoi-ddcutsch. WcaldfiiUililuiii;.. p. '.», |)1. viii, fig. 1. 184',). Pt'copteiis Ilutfoni (Dunk.) Brongii. : Tableau, p. 107. 1869. Ahfhopferis Iluffoni (Dunk.) Schinip.: Pal. Veg., Vol. I. p. oTO [hy lyp. error A. Murchisoni]. 1874. AhtJiopfcrls Ibifton't (Dunk.) Sehini]).: Op. cit.. Atlas, p. 14, pi. .\.\xi, fig. 10. 1888. Aspidhun OerKtaVi Ileer. Lesquereu.x : Proc. I'. S. Nat. Mus., Vol. XI, p. ;32 in part, cpioail Cat. I'. S. Nat. Mu.s.. No. 'IVM. Lcsciuereux's Nos. i)lS, 920, 926, 927. 1888. Asplnrium Farstn-i Deb. & Ett. ]je.s(|uereii.\: Loc. cit. Five .specimens of a large fern were collected by Mr. Woolfe. They are rather poorly preserved and do not show the full character of the plant. Four of these determined by Lesquereux as Aspidiuyn Oersted i Heer and one as Asplcniu))) Fivrsteri Deb. & Ett. The narrower pinnules, occiu'ring on one of the imprints, are not unlike some of those of Aspidiiii/i Oerstedi. The collection of Mr. Dumars contains several large slabs of rock, which show a ntimber of imprints of this form better preserved and more complete than those of Mr. Woolfe. They are also mostly from different parts of the plant. These show that the plant is a fern that can not well l)e distinguished from the fossil that Dunker descrilied from the Wealden of Hanover, with the name Xeuropteris Huttoui. This seems to have been the only specimen ever found. As it is only a small frag- ment it can not show the full character of the plant. Schenk examined this specimen and approved of Schimper's name for it, Alctlioiiteris Hnttoni." He gives a figure of it differing somewhat from that of Dunker, and pi-obably a more nearly correct one. Dimker figiu-es the plant as having two complete pinnae attached to a large rachis, portions of which show the original width. Schenk's figure shows the principal rachis with all of the inai'gin on one side wanting, so that the true width is not seen. The only attached pinna is the upper one, and the end of that is wanting. The next lower pinna has the end preserved, but its full length is probably not given, as the base is defective and the pinna visible is only a ))ortion "Foss. Flor. (liT .Vordncstdciitsch. Wcaldeufonnation (Palaeontographica, Vol, XIX, 1S7I;, p. 217 [1.5], pi. xxix |viii], figs. 1, la. MDN M.VIII 0.5 11 162 :mesozoic floras of united states. of the original one. The specimens in the collections from Alaska show that pinnsp in the position on the frond shown in Dunker's specimen are longer then these, although they are remarkably short for foliage of such size. This fern has an uncommon aspect, and it resembles so closely the specimens from Alaska that there can be no doubt that they l;)elong to the same species. The general aspect is somewhat like that of Thinnfeldia. The specimens collected Ijy Mr. Dumars show that the plant must have been of large size and that it was probably arborescent. One of Dumars's imprints, with much of the lower and upper parts wanting, shows a compound pinna 32 cm. long, with a rachis of the maximum width of 5 mm. This is probably a fragment of a pinna and not of the frond. The specimens are not very well preserved, as the rock is unfavor- alile. ijeing a sandstone with little tendency to cleave. The pinnules are generally a good deal distorted and fragmentary, so that their normal character can be made out only by a careful examination of all the parts and by taking many pinnules. The large fragment represented in PI. XLI, Fig. 1, is probably a compound pinna from well down on the frond. The plant is somewhat different in aspect, according to the position on the frond of the parts. The following may be given as its character: The plant was probably arborescent and of large size. At least a tripinnate character is indicated. The pinnse of superior order must have had a very considerable length, as indicated in the fragment 32 cm. long, which was probably originally twice as long. This is probably only a penultimate pinna. The ultimate pinnae in middle portions of the frond, as shown in PI. XLI, Fig. 1, are short in proportion to the size of the pinnules. Xone are shown entire, but a few are almost so. The largest of these is 85 mm. long. They go off ol)liquely from the main rachis and curve slightly away from it. The pinnules were thick in texture and apparently leathery. The largest pinnules, in l^asal portions of the lower pinnae, are about 2 cm. long and 9 mm. wide in the widest part. They diminish in width and length toward the ends of the ultimate pinnae and in terminal portions of the compound pinnae. Some of the pinnules of larger size are 25 mm. long and only cS mm. wide, but this variation is probably due to distortion from pressiu'e. The distortion and mutilation of the pinnules seem universal, and it is impossible to find a single one not affected. Hence the specimens, when drawn as they now PLANTS FROM ALASKA. 1^V^ are, give a false idea of their original shape and tiiie appearance. The magnified figures of the pinnules and lobes, PI. XLI, Figs. 4. ."), are slightly restored, so as to give the undistorted forms. In the huge penul- timate pinnse, as given in PI. XLI, Fig. 1. which i^rohahly belong to the middle portion of the frond, the true shape of the larger pinnules is oxate to ovate olilong. They are slightly falcate and have lancet-shajjed, subacute to acute tips. They are set obliquely on the rachis and point shghtly fonvard toward the ends of the ultimate pinnte. They are united at base by a decurrence of their dorsal bases. The lower pinnul(\s of lowei' ultimate pinna^ are least united. Toward the ends of tlie ultimate pinna' and in the terminal portions of the compound ones they are more and more united and pass into lobes and teeth, the size being at the same time diminished. They are shown enlarged in PI. XLI. Figs. 2, 8. PI. XLI, Fig. 5, shows, slightly magnified and restored, a portion of an upper ultimate pinna where the })innules are more united and reduced to lobes. Fig. 4. also slightly magnified and restored, gives the true shape of one of the larger pinnules. -""i] The midnerve goes off very obliquely, and at about two-thirds of the distance to the end of the pinnule splits up into t)ranches after the fashion of Cladophlebis, so that the plant is a well-marked type of that genus, and in the absence of fructification must l)e placed in it. The latei-al nerves, in proportion to the size of the pinnules, are quite slender. They are immersed in the leaf substance and are not conspicuous. They go off verv ol)liquely and are forked one or more times. The lowest are the most copiously l)ranched. The forking is notal)ly low down on the nerve, so that the branches are unusually long. On the lower side of the base of the pinnules one or more lateral nerves go oft" from the main rachis. In the more separated pinnules the lower lateral nerves curve away from the midnerve, but in the lobed and dentate forms the interior basal ones often curve inward toward it. This description applies to the large compountl pinna' found by ]\Ir. Dumars, which probably come from the middle portion of the frond. Mr. Woolfe found two rock fi-agments, the ones examined by Lesquereux, that show parts that probably belong to different positions on the frond. One of them, given in PI. XLII, Fig. 1, is apparently a more terminal portion of a principal pinna, which, lower down, would cany as subor- 164 MESOZOIC FLORAS OF UNITED STATES. dinate piniuo the laro;e pininr such as are ^iveii hi PI. XLI, Fig. 1. On this fragment the iiUimate piiiii:e of the form given in PI. XLI, Fig. 1, have been reckiced to pinnules. These differ somewhat in shape from the pinnules lower down. They are narrower in proportion to their length. PI. XLH. Fig. 3. shows one slightly restored and magnified. They are all much mutilated and distorted, so that it is difficult to make out their exact original shape. PI. XLII, Hg. 2, shows an attached pinnule enlarged two diameters. They were prohalily wider than they now appear to be. The general character, however, of tliese pinnules is similar to that of those lower down. The ultimate pinna' of this fragment are quite remote. There are small pinnules, one at least, on the main rachis Ijetween the points of attachment of the rachises of the ultimate pinna', l)ut owing to the imperfect preservation they are not distinctly shown. Another rock fragment found by Mr. Woolfe contains two imprints which probably l)elong to still other portions of the frond. One of these, given in PI. XLIII, Fig. 1, has its pinnules much distorted in shape. Still they are narrower and probably were more acute than the pinnules .shown on PI. XLII, Fig. 1. The fragment is probably a part of the com- pound pinna nearer its termination than are any of the other specimens. This fragment shows a small piece of a penultimate rachis, which carries several fragments of ultimate pinnae, with a numljer of pinnules. Pinnules on this, too, are borne on the main rachis between the ultimate pinnae. PI. XLIII. Fig. 2. shows several attached pinnides enlarged two diameters, and Fig. 3 gives, slightly restored and enlarged, one of the pinnules of this specimen. In contact with this, but not organically connected with it, is the other imprint. This apparently comes from a position on the frond lower than that of any specimen found. It is a fragment of a penultimate pinna, which carries portions of several ultimate ones. These show, with much mutilation and distortion, a numl>er of pinnules that have teeth similar to those on Dunker's plant. The pinnules, however, are somewhat larger than those of that plant and show teeth on both their margins. PI. XLIII, Fig. 4, represents this imprint; Fig. 5 shows a piiHiule enlarged two diameters, and Fig. 6 gives a pinnule of it slightly restored and enlarged. On the same rock fragment there is another specimen of this plant, on a layer deeper in the stone and only partially PLANTS FKOM ALASKA. 165 exposed. Il has pinnules larfici' than those shown in PI. XLIII, Fitr. 4, and more deei)l>' incised. It evidently eoines fi'om a jxtsition still lower on the frond. To judge from the numbei' of specimens of lliis fern fouiul in tlie small collections, it was one of the most abundant and characteristic plants of the Hora of its time. That it was ])retty widely distrihvited is shown by the fact that two specimens of it occui' in the few fossils collected by ]\lr. Schrader, at a locality 180 miles distant fiom the places where Messrs. W'oolfe and Dumars obtained their fossils. The specimens of Mr. Schrader show several fragments of pinnules that are rather deeply incised into lobes, and also dentate ones. PI. XLIII, Fig. 7, represents the specimen in Mi'. Woolfe's collection which Professor Lesquereux referred to Asplciiiuni Fa-rstcri Del). & FAX. Phylum SI^EI^:SIA^TC)3^HYT^V. Class GYMNOSPEKMJ]. Order CYCADALES. Family CYCADACE.E. Genus PODOZAMITES Friedridi Braun. PODOZAMITES DISTANTINERVIS FoUtaille. 1888. Podozamites latipennis Heer. Ije.sciui'roux: Proc. U. S. Nat. Mu.s., Vol. XI, p. -SI, ]>1. xvi, figs. 2, 3. 1888. Zamifes alaskana Lx.: Op. cit., p. .3'2, pL x, fig. 4. 1889. Podozamites distaniinerris Font.: Potomac Flora (Monogr. U. S. Geol. Surv., Vol. XV), p. 179, pi. Lxxix, fig. 5; pi. Ixxxii, fig. 4: pi. Ixxxiii, figs. 1. 2. 6, 7; pi. Ixxxiv, figs. 1, 2, 8, 10, 14, 1.5; pi. Ixxxv, figs. 12, 16. 1902. Podozamites distantineruis Font. Schrader: Bull. Geol. Soc. Am., Vol. XIII, p. 245. Several detached leaflets occur in the collections, mostly in the spec- imens collected by Mr. Woolfe, that seem to be a Podozamites, identical with the form described from the Potomac formation with the name P. distantinervis. Two fragments of it occur also among the specimens collected by Mr. Schrader. Professor Lesquereux identifies most of these specimens with P. latipennis Heer, a form in which the leaflets do not terminate at their 166 iMESO/OIC FLORAS OF IMTKI) STATFS. bases like those, and hnvv a diiTerent shape. One specimen which, in my ()j)ini()n. lielongs to this species, Lesquereux determines as Zamites alaskana Lx., a new species. This specimen, jjiven in his paper, pL x, fig. 4, has, it is true, a .sinus at the base of the leaflets resembling that of some Zamites, but it seems to be an accidental shape due to the rupture of the lamina of the leaflet from the summit of the petlicel. Only the lower part of this leaflet is preserved. It is widei- than mo.st of the leaflets of this species, as in the part ])reserved it attains a width of 24 mm. It may really be a fragment of Podozamitefi grand if oliiis Font. Most of the leaflets of this plant are fragments, but one of the imprints, determined by Lesquereux, is entire and lies near another that is nearly entire. These are shown in Professor Lesquereaux's paper, pi. x\i, fig. 2. They have a maximum width of IS mm., which is maintained to near the ba.se, where it is narrowed rather abruptly and rounded off with an elliptical shape. The entire leaf narrows gradually near its free end and terminates in a lancet-shaped tip. It is 105 cm. long. The two leaves are slightly curved and have a somewhat ensiform shape. The character seen in these two seems to be found in all the leaflets, for they dift'er only in length and in proportion of width to length. None show any portion of the pedicel except one of the fragments among Mr. Schrader's specimens. This is too poorly preserved to show more than the fact that it is a portion of the pedicel. Fig. 3 of the same plate gives a portion of a leaflet occur- ring on the same rock fragment that carries the leaflets given in fig. 2. This shows very perfectly the terminal portion of the form. It seems to be a leaflet somewhat shorter than the leaflets in fig. 2, but there is no great difference in the length of any of the leaflets. I have identified this plant with Podozamites distajdinems, a Lower Cretaceous fossil, but it should be stated that, except in dimensions, the leaflets resemble some forms of P. lanceolatus Eichwaldi (Schimp.) Heer and P. lanceolatus latifolius (Fr. Br.) Heer" that Heer has described from the Jurassic of Siberia. They may especialh' be compared with the figures given on pi. xxvi. These are, however, decidedly smaller than the leaflets of the Alaskan fossil, with the possible exception of fig. 6. This seems to be a fragment of a large leaflet, which may not be P. lanceolatus latifolius (Fr. Br.) Heer. " Fl I'"oss. Arct., Vol. IV, Pt. II (Beitriige zur Jura-FI. Ostsibiriens und d. Amurlandes), p. 109, pi. xxvi, (ifrs. .5, (>, 8b, c. PLANTS FROM ALASKA. ' 167 PODOZAMITES GKAXDIFOI.US FoiltaioO? PI. XLIV, Fig. 1. ISSS. Bainii pahndta Ilcor. Lcsciuoreux: Proc. U. S. Nat. Mus.. Vul, XT. ]>. ;;i, 1>I. \vi. lijr. 4 (<|U(>!i(l Cat. V. S. Nat. Mus., No. 2437, Lcs(|iicr(Hi.\'s Xos. !)1(), !»lla.) 18811. Podozamites (/rami if alius Font.: I'otoinac Flora (Monoj^n'. V. S. Geol. Surv., Vol. XV), p. 180, pi. Ixxxii. fig.s. 2. 2a: pi. Ixxxiii. fii:. .">. In the collection made liy Mr. Woolfe there are several specimens showing fragments of what must liave been a very large leaflet. Lesque- reux identified most of these with Baiera palmata Heer, a .Ivu'assic plant. He seems to have been induced to make this determination b>- the fact that one of the leaflets shows a .split at one end, extending a little way down toward the l)ase. Only one of the impressions shows this. A careful examination of it indicates that this is an accidental fission and that the forms before mutilation were entire. Lesquereux's figiu-e is therefore erroneous in tliis particular. The plant is quite problematic and I am not sure of its true character. The leaflets are too imperfect for a correct determination. They appear to be nearer Podozamites than an>- otlier fossil. One of the specimens shows a terminal part. This, given in PI. XI.R', Fig. 1, must have been a very large leaf. The end shown in it seems to Ije the free end or termination of the leaflet. At the opposite end it is mutilated and much of the length is lacking. Still, it shows a length of 9 cm. and a wddth at the broken end of 3 cm., allowing for the splitting which occurs here. The specimen figured by Lesquereux, although imperfect at both ends, shows a length of 11 cm., with one margin nearly entire. This specimen shows conclusively that the plant is not ;i Baiera, but indicates strongly that it is a leaflet of the form of Podozamites. This margin is slightly ciu'ved and indicates that the leaflet ma>' have been ensiform. The only forms resembling this plant are those of Podozamites gmndifolius of the Potomac l)eds. The size of the leaflets and the form, so far as it is indicated in this specimen, are strongly suggestive of the Potomac fossil. The nerves are not well preserved, but so far as they can be made out they agree well with those of the Potomac plant. They seem to be strong and flat, appai-ently made up of two nerve strands. On the same piece of rock and partly overlapping this specimen is the impression of a leaf of Nageiopsis longifolia Font., to be mentioned below. 168 MESOZOIC FLORAS OF UNITED STATES. Order GINKGOALES. Family GIXKGOACE.E. Genus BAIERA Friediich Braun. Baiera gracilis (Bean) Bunbiuy. PI. XLIV. Fig. 2. 1S4.3. Schizopteris gracilis Bean in Morris: Cat. Brit. Foss.. p. 20 (from Bean's manu- script). 1S40. Baiera sp. Brongn.: Tableau, p. .38 (fide Bunbury: Quart. Journ. Geol. Soe. London, Vol. VIL p. 182). 18.5L Baiera gracilis (Bean) Bunbury: Quart. Journ. Geol. Soc. London. Vol. \'n, p. 182, pi. xii, fig. .3. 1865. Cycloptetis gracilis (Bean) Zign.: Osserv. sidle Felci Foss. dell'Oolite. p. 22. 1878. Schizopteris digiiata 'Willn. [non (Brongn.) Gein.] in Saporta: Plantes Juras- siques. Vol. Ill, pp. 277. 279. 1902. Baiera gracilis (Bean) Bunb. Schrader: Bidl. Geol. Soc. Am.. Vol. XIIL p. 24.5. In the collection made b}- Mr. Schrader there is a single specimen that agrees exactly with Baiera gracilis (Bean) Bunb. It shows a por- tion of a long petiole which bears at its summit two equal segments. These divide dichotomously into several narrow laminse, which are presers-ed for only a portion of their length. The .specimen shows espe- cially a striking resemlilance to the figure of this Baiera given by Seward in his Jurassic Flora of the Yorkshire Coast, pi. ix, fig. 5. It resembles B. Muensteriana (Presl) Heer of the Rhetic formation, and Seward points out the resemblance of some of the forms of B. gracilis to that plant. Genus GIXKGODIL'il Yokoyama. GiNKGODiUM ? ALASKEXSE Fontaine. Pi. XLIV, Figs. 3, 4. 1888. Baiera palmata Heer. Lesquereux: Proc. U. S. Nat. Mus., Vol. XI, p. 31, pi. xvi. fig. 5. Two fragments of peculiar leaves occur in the collections, one in the collection of Mr. Woolfe and one in that of Mr. Dumars. The speci- men found by Mr. Woolfe was determined by Lesquereux as Baiera PLANTS FROM ALASKA. 169 palmata Heer and figured in his paper. pL xvi, fig. 5. It. however, shows no trace of a division of the lamina of the leaf. The other, given in PI. XLIV. Fig. 3. was obtained by Mr. Dumars. Both give the basal parts with a portion of the petiole. These leaves lack their temiinal parts. They narrow gradually to a wedge-shaped base and expand to a fan shape in the opposite du'ection. They show no division of the lamina in the parts preserved, but may higher up have Ijeen palmately divided. The base, in its prolonged wedge shape, differs from the Ginkgo leaves associated with these forms, but the principal difference is in the nerves. These are shown with some distinctness in PI. XLH'. Fig. 3. Fig. 4 shows the nature of the nerves. The margins are thick- ened to forni a parent nerve that sends ofi". veiy ol)lic[uely, secondary- nen^es that enter the lamina of the leaf. They fork at then* insertions and then are mostly single. Occasionally at long inter\-als a second forking occui"s. but this is ^•ely rare. The branches are approximately parallel, diverging slightly in ascending in the leaf. They are quite strong. The nerves of the central part of the leaf ascend from the top of the petiole. In many respects these leaves are like the genus Ginkgo- dium, established l^y Yokoyama for certain forms from the Jurassic of Japan." But the Japanese plants have slender nerves that do not fork at all. They go off from the marginal nerve straight to the summit of the leaf and are parallel to the axis of the leaf. As, however. Yoko- yama found only one species, it is possible that the limits of variation of the genus may include the Alaskan species. This fossil resembles also the forms described from the Permian of southwest Pennsylvania and West Mrginia as Saportsea.*' This has the marginal nerves and the branching lateral ones, but the difterences are too gi'eat to pennit these leaves to be placed in that genus. The material is so imperfect and small in amount that the Alaskan fo.ssil can not be positiveh' identified with Ginkgodium. " Jurassic plants from Kaga.etc: Jour. Coll. Sci. Imp. Univ. Japan. Vol. Ill, Pt. 1. pp. .56-58. pi. ii.fig 4e: pi. iii, fig. 7: pi. viii; pi. i.x. figs. 1-10. 10a: pi. xii, figs. 14. 1.5. 6 The Permian or Upper Carboniferous flora of West Virginia, by Wm.M. Fontaine and I. C. White: Second Geol. Survey Pennsylvania. Report of Progress, PP. ISSO, pp. 99-103, pi. x.^ixriii, figs. 1-1. 170 MESOZOIC FLORAS OF UNITED STATES. Genus GINKGO Kaempfer. Ginkgo digitata (Brongniart) Heer." PI. XLIV, Figs. 5. 6. Several specimens of a Ginkgo occur in the collection of Mr. Dumars, which in general character agree well enough with Ginkgo digitata to be placed in this species. The leaf represented in PI. XLIV, Fig. 5, probabh- had a somewhat different form from that now shown. It is mutilated at the summit and slightly distorted at base. The free ends of the lol)es appear truncated, but this may be due to the removal of the tips in splitting the rock. This leaf has four divisions and evi- dently had originally no more. They are irregular in wichh, and three of them are considerably wider than the lobes of the leaf collected by Mr. Woolfe and figured by Lesquereux in his paper, pi. xvi, fig. 6. This leaf is also less deeply divided and a portion of a stout petiole is shown. Another specimen, more imperfect than these, gives a transition from the one to the other. This is shown in PI. XLIV, Fig. 6. Ginkgo Huttoxi (Sternberg) Heer.'' 1888. Ginkgo muUinenns Heer. Lesquereux: Proc. U. S. Nat. Mus.. \o\. XI, p. 31, pi. xvi, fig. 6. One specimen probably belonging to this species occurs in Mr. Woolfe's collection. Lesquereux identified it with Ginkgo multinervis Heer from the Cenomanian beds of Atane, Greenland. It has three almost entire partitions, with a fragment of a fourth on the right side. It is, however, obviously incomplete on this side and had at least one more lobe. The leaf is divided almost to the base, and the divisions are obtuse at the free ends, elliptical in form, and narrow to a wedge shape at the base. The dimensions of this specimen agree very well with those of G. Huttoni. Ginkgo Huttoni magnifolia Fontaine?'^ PI. XLIV, Figs. 7, 8. In Mr. Dumars's collection there are several imprints of fragments of very large Ginkgo leaves that must have much surpassed in size any " For tlif full synonymy of this species see pp. 121-122.— L. F. W. *' For synonymy of this species see p. 12.3. — L. F. W. cSee p. 124. PLANTS FROM ALASKA. 171 known leaveo of Ginkgo (ligitaUi. They closely resemble a large Ginkgo leaf found in the Jurassic (Lower Oolite) flora of Douglas County, Oreg. This form has been described by me as G. Hnttoni magnifolia (supra, p. 171). The leaves now in question resem])le the Oregon plant in their great size, in the irregular width of the tlivisions of the leaf, in the remote- ness of the nerves, and in their great strength. The size of some of these Alaskan leaves makes it improbable that they belong to G. digitata. One of the fragments, which has nuich of the summit of the leaf missing, is still 5 cm. long, while a spread of 7 cm. is shown on one very imperfect specimen. The nerves of these are more remote than those of G. (ligitata and much stronger. PI. XLIV, Fig. 7, gives a fragment of one of these leaves which is apparently divided into only two very wide lobes. But even this is not certainly an original partition of the leaf. It has much the appear- ance of an accidental division. Some idea of the size of the leaf may be oljtained from it. Another specimen is given in Fig. 8. In this the divisions are clearly accidental. In it a portion of a stout petiole is shown. As, however, the material is very imperfect, the true place of the forms can not l)e positively determined. Order FINALES. Family TAXACE.E. Genus NAGEIOPSIS Fontaine. Nageiop.sis LONGiFoLiA Foiitaiue. PI. XLV, Figs. 1-.5. 1887. Irites alasTcana Lx.: Proc. U. S. Nat. Miis., Vol. X, p. 36." 1888. Baiera palmata Heer. Lesquereux: Op. cit.. Vol. XL p. 31 in part, quoad Cat. U. S. Nat. Mus., No. 2437, Lesquereux's No. gil.'' "The four specimens thus named by Professor Lesquereu.x were collected bv Mr. Woolfe but were either received in advance of the main collection or otherwise became separated from it and were sent to Lesquereux and described by him a year earlier than the re.st. They were overlooked in sending the collections to Professor Fontaine and not included in his report. They were subsequently sent to him, and in his letter dated March 17. lt)02,he says of them: '"The specimens sent last do not call for any modification of my report, as they are all Xageiopsis Uintj'ifolin." The best specimei\ is shown in PI. XLV, Fig. h. — L. F. W. '' .\lthougli Professor Fontaine .says that Lesquereux did not mention the specimens of this species in Mr Woolfe's collection, nevertheless he labeled the one having Lesquereux's No. 911 -Vayoo/wis Innijifolia. This was one of tho.se that Lescjuercux referred to Bdiera jmhnaUi Heer. It is here represented in PI. XLV, Fig. 2. — L. F. W 172 MESOZOIC FLOKA.S OF UNITED STATES. 1889. Xageiopsis longifoJia Font.: Potomac Flora (Monogr. U. S. Geol. Surv.. Vol. XV), p. 19.5, pi. Ixxv, figs. 1, la, lb; pi. Ixxvi, figs. 2-6; pi. Ixxvii. figs. 1, 2; pl. Ixxviii, figs. 1-.5: pi. Ixxix, fig. 7; pi. Ixxxv, figs. 1. 2, 8, 9. 1902. Xageiop.nslongifolia Font. Schrader: Bull. Geol. Soc. Am., Vol. XIII, p. 24.5. A considerable iiuml^er of fragments of detached, narrow, strap- .shaped leaflets occur in all the collections. All of those in Mr. Schrader's collection are found on a single rock fragment. Lesquereux does not seem to have noticed those in Mr. Woolfe's collection. At least he gives no description of them. He may have regarded them as forms of Podozamites latipennis Heer, but they are quite different from some of the fossils he identified with this species. They are uniformly narrow and for most of the leiigth of the fragments do not differ in width, so that the margins are parallel. They also narrow more at base than the true P. latipennis, and seem to have been attached l:)y a short pedicel, which is not the case with Heer's plant. None of these leaflets are com- plete. They all lack the terminal parts. As none of them are attached, and in only one case is the basal end visible, it is not possible to determine them positively. The fragment occurring on one of the rock specimens obtained by Mr. Dumars shows a length of 10 cm., with both the Ijasal and the ter- minal portions lacking. The average width is II mm., which it main- tains to near one end, where it is narrowed to 9 mm. This is apparently the basal end, but a portion is still lacking here. This is given in PI. XLV, Fig. 1. The nerves are rather remote. They are parallel and, as shown in one of the imprints in which the base is preserved, fork only on leaving the pedicel. ^lost of the imprints are shorter fragments in which the margins of the leaflets and the nerves are parallel. The width of the leaflet given in Fig. 1 seems to be the average one, Ijut there are narrower ones, having a width of 7 mm., and wider ones, with a width of 14 mm. One of these larger leaves is shown in Fig. 2. This has a length of 95 mm. It narrows at one end to 7 mm., after the fashion of the one shown in Fig. 1 . This represents the specimen referred by Lesquereux to Bniern palmata Heer. One of the smaller leaves, with a maximum width of 7 mm., is shown in Fig. 3. This occurs on one of the rock fragments obtained by Mr. Schrader. It shows the base of the leaflet and the mode of narrowing and rounding off there, but the pedicel is lacking. This sp^ci'^^'i p-i-v^s the l)ase better preserved PLANTS FROM ALASKA. 1"3 than any other. It shows also the nerves very distinctly. Fig. 4 repre- sents the basal portion of the principal impression enlargefl two diameters. In the mutilated condition of the leaflets and their detached character it is not possible positively to determine the place of this plant. It agrees well with the conifer Xageiopsis longifolia," first found in the Potomac formation. Some of the leaves are rather wider, but the ditl'erence is not gi'eat. The widest of the Potomac leaves are 12 mm. wide. The leaflets agree especially well with those of figs. 2 and 5 on pi. Ixxvi of the work just cited. The nerves also agree well with those of the Potomac plant. They are sharply defined and rather remote. They are parallel and simple to the base of the leaflets. At the base, on entering the pedicel, they fork once and converge l)y curving sharply. The considerable numl^er of fragments shows that the plant was not uncommon. CONCLUSIONS. In the outset it should be stated that the amount of material in tliese collections is quite small. Much of that obtained is very imperfect. A large portion of it is composed of two or three species, and there is a lack of decisive fossils. Hence it would be understood that any conclu- sions that may be drawn will be sul)ject to doubt. The following is the list of plants which have any significance: L Dicksoiiia Sapoitana lleer. One specimen. 2. Onychiopsis |)silotoides (Stokes & Webb) Ward. Several specimens. 3. Cladoplilchis vaccensis Ward. One specimen. 4. Cladophlebis alata Font. Many specimens in proportion. .5. ClafTophlebis Huttoni (Dunk.) Font. Proportionally many sj)ecimens. 6. Podozamites distantinervis Font. Several specimens. 7. Podozamites grandilolius Font.? Several specimens. 5. Baiera gracilis (Bean) Bunb. One specimen. 9. Ginkgodium ? alaskense Font. Two specimens. 10. Ginkgo (ligitata (Brongn.) Heer. Several specimens. 11. Ginkgo Huttoni (.Sternb.) Heer. One specimen. 12. Ginkgo Huttoni magnifolia Font.? Several specimens. 13. Nageiopsis longifolia Font. Proportionally a good many. These thirteen forms are by no means equal in value for determining age. They differ much in the number of specimens and in the definiteness " Potoiiiac or Younger Mesozoic flora: Monogr. V. S. Geol. Surv., Vol. XV, pp. 19.5, 196, pi. Ix.w, fig. 1; pi. Ix.xvi. figs. 2-6; pi. Ixxvii, fig.s. 1, 2: pi. Ixxviii, figs. 1-.5: pi. Ixxix, fig. 7; pi. xxxv, figs. 1, 2, 8, 9. 174 :\IESOZOI(;' FLORAS OF UNITED STATES. of their (lotorniination. To have value in determining age the mere presence of the species is not all tliat is required. The plant must be abundant in the flora and characteristic of it. It may be a survivor from an oldei- flora. The proportion of specimens in a collection is the only feature that gives a hint on these points. It of course must not be insisted on too strongly, for there are other conditions besides the actual relative abundance that may give a large proportion of the specimens to one species. Cladophlebis vaccensis, Dicksonia Saixirknia, Ginkgo digitata, G. Huttoni rnagnifolia'^., and Baiera gracilis are notably fossils of the Lower Oolite. Ginkgodinml alaskense is a new species, and if it be a true Ginkgodium, its nearest kin is found only in the same formation. On the other hand, the following belong to the Lower Cretaceous, taking the Wealden as belonging to that formation: Cladophlebis alata, C. Huttoni, Onychiopsis psilotoides, Podozamites distantinervis, P. grandi- folius'!, Xageio))sis longifolia. These plants are, as stated, not of equal value in determining age. Cladophlebis vaccensis, Dicksonia Saportana, and Baiera gracilis have each only one specimen. Hence we may conclude that they were not abundant in the flora, and they may be sui'vi\'ors from an older one. Podozamites grandif alius"] and Ginkgo Huttoni magnifoliai are not posi- tively determined. Leaving these and the probable Ginkgodium out of the question, we have, as the fossils of most value for fixing the age, five plants, viz, Cladophlebis (data, Onychiopsis psilotoides, Cladophlebis Huttoni Podozamites distantinervis. Ginkgo digitata, a very small list. The Cladophlebis psilotoides shows some difi"erences from most of the described Lower Cretaceous forms, which somewhat impair its value as evidence. Cladophlebis alata and C. Huttoni, if we may judge from the mimber of their specimens, must have l)een abundant and highly characteristic of the Alaskan flora of their time. All of the more im- portant plants except Ginkgo digitata are Lower Cretaceous, and if we take simply their percentage in the flora the evidence is ovei-whelming in favor of the Lower Cretaceous. I am, ho^\•e^'er, inclined to attach great weight to the considerable proportion of Ginkgos of Jurassic type. The Ginkgos in Lower Oolitic times were immensely developed in the Amur region in Siberia and in the northwestern part of the United States, PLANTS FROM ALASKA. 175 showing a remarkable resemblance in their forms. The resemblance is such as to suggest land coiuiection in the lower Oolite epoch between Asia and North America and a common flora. In Lower Cretaceous times, however, they had become nearly extinct on the American continent, even in those parts where they had been so prominent in the Lower Oolite. In all the Lower Cretaceous floi'a of the northwestern region the only Ginkgo found occurs in the Kootanie beds of Canada. Sii' William Dawson, in his Mesozoir Floras of the Rocky Mountain Region of Canada," describes three fossils which he regai'ds as Ginkgos. One of these, given on pi. ii, fig. 1 , he identifies with (I. sihin'ra Hear; another, shown on pi. ii, fig. 2, he determines as G. Icpiiln Heer; and a third, depicted on pi. ii, fig. 3, he names G. nana. His (/. Ic/iida and G. nana are evidently not Ginkgos, but are probal)ly a Baiera, and l)oth the same species. His Ginkgo sibirica is a true Ginkgo and may be that spocies. It is the sole survivor apparently and is much smaller than most of the types common in the Lower Oolite. \i\ the Alaskan fossils the relative ai)undance of the specimens of the two Ginkgos shows this type of plant is still present in force. The large size of the leaves shows that it is probal)ly still in full vigor. This indicates that the time in which these Ginkgos li\-ed in Alaska is not so late as the Lower Cretaceous. The characteristic Lower Oolitic forms still constitute a large element in the flora. The floras of the dift'erent parts of the northern hemisphere in the Lower Oolite and the Wealden are compai'atively well known; the vegetation of the intervening time, especially in its foliage, is little known. It is probable that many of the Jurassic types found in the Wealden or Lower Cretaceous flourished in the Upper Oolite and survived in the Lower Cretaceous. The finding, then, in a locality of such Lower Cretaceous forms asCla(loj)hlebis alata, Onychiopsis psilotoides, Cladophlehis Huttoni, etc., does not necessarily prove a Lower Cretaceous age. The age of the formation yielding the Alaskan fossils, as indicated by them, is not older than the Lower Oolite, and not younger than the Lower Cretaceous, but is probably between them. "Trans. Roy. Soc. Canada, Sect. IV, Vol. Ill, 188.5, p. 8. 176 MESOZOIC FLORAS OF UNITED STATES. ;-. PLAXTS FROM XORTHRRX CALIFORXIA. [Under date of Deceml)er 26, 190U Professor Fontaine reported as follows upon the collection made by Mr. Storrs in Trinity County, Cal. — L. I". W.] I have examined the specimens collected by Storrs from California, 2 miles northwest of Slatonis. Most of them are shale frajjments, with distinct cleavage, and all have a more or less pronounced cleavage. Nearly all of the imprints have suffered so much from maceration that they show no character. They have evidently- drifted far from their place of growth. BKAciiYPHYLLrM? Storrsii Waiil n. sp." PI. XLV, Fig. 6. 1903. BrachyphyJIum ? sp. Font, in Dillcr: Am. Joiirn. Soi., 4th ser., Vol. XV, p. 352. The best preserved imprints are cones which, in some cases, preserve enough of the plant tissue to give an idea of their nature. The best preserved of them lun'e been much compressed and distorted by pressure, which has caused a creep of the plant substance along the planes of cleavage, so that it is impossible to determine positively even their generic character. Besides the cones there are some poorly preserved bits of twigs which probably belong to the same plant as the cones. There is a great difference in the size of the cones. Some of them seem to be mature, some imma- ture, and there appear to be some male strobiles. All the imi)rints which show any recognizable features are strongly suggestive of Brachyphyllum. They may, however, belong to the genus which I described as Pagiophyl- Jum duhium from the Comanche of Texas,'' which Nathorst makes a new genus, Pseudofrenelopsis.'' If the cones belong to Pseudofrenelopsis the species is probably different from the Comanche form, for the cones now in question are decidedly larger. Their true size, however, is not given in their present form, for they are mostly broadened by the creep of the shale. The same creep has greatly distorted the form of the cone scales. « As Professor Fontaine does not assign to this plant any specific name, and as it is likely to be the suhject of future di.scussion, I propo.se for it the name lirachyphynnm ? Storrsii, for the collector. — L. F. W. '' Notes on .some fo.ssil plants from the Trinity division of the Oonianche series of Texas: Proe. U. S. Nat_ Mus., Vol. XVI, pp. 2(il-2S2. See p. 271, pi. xxxix, fif,"*- - ' ' ''Beitr. zur Geologic und PalaDntolofric dcr Rcpiililik Nh-xiro. von .1. FcMx und H. Lenk, Leipzit;, 1893, II. Theil, 1. Heft, pp. 51-.54. I'LANTS FKOM CALIFORNIA AND MONTANA. 177 Some few of these are shown with Htlh' (hstortioii.'aiHl tliey resemble those of Braclivphylhim; that is, they are thick and rhoml^ie in form, with tlie greater (hniensions transverse to tlie axis of the cone. Most of them, however, are in this trans\'erse cHrection so much elongated and distorted that lliey appear as parallel i-aised lines. The mature cones may he compared with the cone h. Marc (UKn/uni Brongn., as given l)v Saporta in Paleontologie Francaise, Pluntes Jurassicjues, Vol. Ill, pi. xxxix, fig. 2. They are, however, broader than that, which may l)e due to the distortion transver.se to the axis of the cones. The cone scales are larger than those of B. Moreauanum, and the species is probably different. There are also, as stated, obscure bits of twigs, which appear to belong to the same plant as that carrying the cones. The>- are mostly decorticated, but a few .show vague traces of leaf scars similar to those of Brachyphyllum. The apparent sti'obiles are small cylindrical fragments with chaffy scales. They probably belong to the plant that shows the larger mature cones. There are several small elliptical to globose cones that are much smaller than the mature ones. They apparently have the same kind of cone scales as the larger cones, l)ut smaller and thiimer. These mav be immatin-e cones of the plant carrying the larger cones, the male strol)iles, and the leafy twigs. Brachyphyllum is most developed in the Jurassic and Lower Creta- ceous. If we may regard this plant as belonging to that genus, then, so far as its evidence goes, the strata are Jurassic or Lower Cretaceous. But as the generic place of the fossil can not be determined positively, and the amount of material is so small, the age can not be certainly- fixed. 6'. PLAXr.s FROM NORTHERN MONTANA. Sequoia Reichenbachi (Geinitz) Heer." PI. XLV, Ficj.s. 7, 8. This specimen was obtained by Dr. A. C. Peale from the east slope of the Bridger Range, north of Bridger Creek, Montana, and is labeled by Doctor Peale as Jurassic. Professor Ward gives as the more exact locality 4 miles northeast of Bozeman, Mont., on the right l)ank of Bridger Creek. The fossil is a small bit of a twig 25 mm. long, with a o For the synonymy of this species see Nineteenth Ann. Rep. U. S. Geol. Surv., Pt. II, 1899, p. 674. MON XLvm — 05 12 178 MESOZOIC FLORAS OF UNITED STATES. number of fairly well-preserved leaves. It is clearly a Sequoia, and is most probably S. Rcichcnhachi, as the leaves have the size, shape, and mode of insertion of those of that species. It indicates that if the for- mation containing it is Jurassic it is the upper part of it. But it may well be Lower Cretaceous. 7. PLANTS FROM SLATE SPRINGS, MONTEREY COAST, CALIFORNIA. Sequoia Fairbaxksi Fontaine n. sp. PI. XLV, Figs. 9-11. This is the plant referred to in Pt. II of the Twentieth Annual Report of the United States Geological Survey, pp. 338, 339. Four specimens were obtained, one of which shows no character. They were collected by Mr. H. W. Fairbanks from Slate Springs, California, in rocks underlying the Knoxville group of the Lower Cretaceous. The specimens are poorly preserved and have suffered from maceration. The form given in PL XLV, Fig. 9, is a fragment of the largest leafy branch that was found. The leaves are poorly preserved and pressed close to the stem, so that thej' can not be seen distinct from it. They seem, however, to have the long slender form that is better shown in the speci- men depicted in Fig. 10. The stem given in Fig. 9 seems to have had a diameter of 5 mm. Fig. 10 represents a much smaller twig. This is 65 mm. long and 15 mm. thick. It carries scattered along its length a num- ber of developed leaves and at its end a number of undeveloped ones, forming a bunch similar to those shown on the small twig depicted in Fig. 11. The developed leaves, as shown in Fig. 11, may, for description, be taken as the normal ones. They are a good deal like those of Sequoia Reichenbachi, and the plant is apparently a sequoia of the Reichcnbacid type. The leaves are 15-20 mm. long. They narrow very gradually to an acute point and widen toward the base. They are decurrent and strongly incurved, showing a slender midrilj. They are more slender and thinner in texture than the leaves of S. Reichenbochi. Fig. 11 gives the terminal portion of a small twig on which the leaves appear to be undeveloped. These leaves are very narrow, short, and straight. They are pressed closely to the stem. JUKASSIC CYCADS FROM WYOMING. 179 The plant is not unlike those from tlio .lurassic, called by Heor Elatides, and may he compared with E. falcala:' l)uf the leaves are largei' than those of that plant. It may be httingly named from its discoverei- Sequoia Fairbanksi. JURASSIC CYCADS FRO^VI AVVOMIXG. Since the appearance of the first paper of this series, in which all the Jurassic cycads from the Freezeout Hills of Wyoming that were known to me at that time were described and figured,'' two aflditional invoices of material from the same restricted Ijed have Ijeen sent to the National Museum l)y Professor Knight under the same conditions as those relatiuL^ to the first invoice. The formei' of tliese invoices consists of the collection made by me on the occasion of my visit to the locality in 1899, an accovmt of which is given in the first paper, but the full treatment of the collection could not 1)6 then made, as it was necessary to go to press with the paper before the collection could be studied (see p. 387 of that paper). As soon as I found time, however, I had the collection unpacked and the speci- mens numbered according to Professor Knight's instructions. The.se were to continue the numbering from the last number of the first invoice as far as the specimens extended. The numl)ering was on the basis of 500, and the first invoice included Xos. .500.1 to 500.87, although these numbers included .several specimens of fossil wood and one Ijone taken from the same bed, the latter not sent with the cycads. Only a few large specimens or nearly complete trunks wei'e found l)y me and the collection consisted chiefly of fragments, many of them quite small, some of them mere chips or splinters. I was careful to save almost everything that could be seen certainly to belong to a cycadean trunk, in the hope that, coming as they did from the same bed, a few of them might be foimd to be the missing parts of incomplete trunks in the first invoice. In this, as will be seen, I was not mistaken, although the result is not so satisfactory as might perhaps have been expected. The number of such small fragments was very large, and when they were all numbered they extended the list from No. 500.88 to 500.687, including therefore, by a curious coincidence, just 600 specimens. a Fl. Foss. Arct., Vol. IV, Pt. II (Jura-Fl. Ostsibiriens), pp. 7f)-,S0, pi. xiv, fig.s 6 (ib 6d (■Twentieth Ann. Rep. U. S. Geol. Surv., Pt. II, 1900, pp. 382-417, pi. lx.x-cl.x.xvii. 180 MESOZUIC FLORAS OF LNFrED STATES. Mention was made in the first paper (p. 387) of a collection that had been made by Mr. Charles Gilmore l)efore my arrival. This, I was told, was stored in a building in Medicine Bow, l)ut as our j^arty hm-ried through that place to reach the field, I did not take time to hunt it up and examine it. It was expected that an effort would be made to inciease the colkM-- tions l)v plowing the ground deepl>' with a subsoil plow, as might easily be done. I had dug out a number of fine trunks with my mattock that were not visil)le from the surface, but this process was slow and lal)orious, and it was thought that such subsoil plowing might reveal many more. At the close of 1902 I learned from Professor Knight that the Uni- versit^y of Wyoming had arranged with the Carnegie Museum to plow the ground on wliich the cycads were found and divide the results between the two institutions, Init that the degree of success was not what had been anticipated. Professor Knight stated that before the ground was plowed he had fovmd ' 'one of the finest specimens that has ever been taken from the place. ' ' This I have not yet seen. He sent me, however, as one of the results of the plowing, a fine terminal bud, which he thought might be a cone. As nearly as I can judge, it belongs to Cycadella jurassica, and at his suggestion I have given it the next number of the museum of the University of Wyoming, first series, which is 500.688. On March 20, 1903, Dr. T. W. Stanton turned over to me a specimen collected by W. T. Lee from the same bed in the Freezeout Hills. It is a small fragment from the side of a large trunk showing half a dozen large scars that indicate that the trunk was that of Cycadella icycnningensis. It is deposited in the National Museum with the locality numl;)er, 3050, of the United States Geological Survey. During the summer of 1901 a third invoice arrived, purporting to contain all the specimens collected to that date. The larger trunks, at least, are doubtless the ones previously collected by Mr. Gilmore, but nearly all the specimens in this invoice are fair-sized fragments, and there are very few small pieces, such as many that I saved. This last collection was numbered before it was shipped and on a different basis. It bears the numbers of the Museinn of the I'niversity of Wyoming from No. 100.201 to No. 100.353, thus containing 153 specimens. There are, therefore, in the additional material to be studied 753 speci- mens, great and small. Yet in all this there are not a dozen trunks that . IlKASSIC CVCADS KKOM \\V()MlN(i. 181 are even approximately coiiiph'tc'. The rest arc f'rajiincnts of all f>izes, l)ut many of them very small, llic least \v('i simultaneously with their study. .\t first it seemed tliat only a very few could be identi- fied wnth described species on account of theii' fragmentary character, but prolonged and mimite in.spection and comparison with the ty])es gradually revealed chai-acters that could not otherwise be detected, and finally eiuibled me to ventui'e a (provisional reference in the case of a large number to the species described in the first paper. The (juestion whether the new material contains any additional species beyond the twenty species previously described can not l)e definitely settled, but I incline to think that some of the fi'agments l)elong to species different from those of the first collection. This, however, is not certain from the scanty material, and I have therefore referred such with doubt lo tlie species that they most closely resemble. It does not seem that any of the complete trunks or large characteristic fragments l^elong to new species, although some of them possess characters not seen in the original types. This is notably the case with those that I am ol)liged to I'efer to species foimded on only one or two specimens, but in such cases this was to be expected. I shall take up the species in the same ordei- in which 1 hey were treated in the first paper and make such additional notes on each as the study of the later material calls for. In view of the probal )ilit y t hal t he specimens will be one day taken up and sul)jected to microscopic study fi'oni the 182 MESOZOIC FLORAS OF UNITED STATES. standpoint of internal structure, I shall append to my notes on each species a list of the nunil^ers of specimens that I provisionally refer to it. This ma}- prove useful as a basis for such work, although I do not doubt that the study of their internal structiu-e will require many changes in the classification. I therefore do this at the risk of liaving many errors — if such they can be called — subsequently corrected, but as all I'efei'ences must l)e regarded as provisional, and as the extremely fragmentary character of the material makes certainty unattainable. I am sure that due allowance will l)e made and that no one will attribute necessaiy changes to carelessness on m}' part. ticmis C'YC'ADELLA Wind. 1!)00. Cycadelhi Ward: Proc. Wash. Acad. Sci., Vol. I, p. 263, pi. xiv: Twentieth Ann. Rep. U. S. Geo). Siirv.. 1898-99, Pt. II, |>. .392, pi. Ixx. All the specimens of the two collections luider consideration belong to the genus Cycadella, and none of them certainly represent species not already described. Cycadella Reedii Ward. PI. XLVII, Fig. 3. 1900. Cycadella Reedii Ward: Proc. Wash. Acad. Sci., Vol. I, p. 264, pi. xv; Twen- tieth Ann. Rep. U. S. Geol. Surv., 1898-99, Pt. II, p. 393, pi. Ixxi-Ixxvi. An imexpectedl>' large number of fragments had to be referred to this species, most of them more or less doubtfully, but in all cases the characters show a closer relationship to it than to any other descril:)ed species. They are mostly fragments and there is no complete trunk. They represent small trunks, usually, so far as can be judged from the parts we have, smaller than any of the original type specimens except No. 500.10 (see pi. Ixxvi). They are generally from tiiuiks larger than that, but reseml)le it in other respects more than they do other types. They show variations in the internal structure which may be specific, but as all the type specimens previously described were so nearly complete and the structure was not shown, it is not known what the interior would reveal. There is, however, nothing in any of the fragments referred to this species that conflicts with the description given. PI. XLVII, Fig. 8, represents one side of the specimen No. 100.239, which is a fragment weighing 0.47 kg., of which the top and the whole of one side are wanting. JURASSIC CYCADS FROM WYOMING. 183 Cycadella Beeciieuiaxa Ward. NonoiJ.he specimens in tl,e secon.l ,„„1 llnv,l invoices were .vfen-ed to tl,is species. The types wee not in „,y lu.n.ls, l-en,^ at the \ ale Un V « tv Mnseun. They consist of No. 128 of that nu.se.nj, and ^a 5,^0 54 ofthe Musettm of the University of Wyom.ng, winch I found to T; comple,«„t of the former, and which, on th,s acconn Pro e.sor Kniti™ gave to Vale to complete the specimen the,.. .41th„ngh th tZ£ tvas carefuUv described and Inlly figured, am I have a d,stu,ct lu Timptession oi its appearance, it is not inM.oss.hle that ,f the spec- irn' 3 been before me I sho.dd have provisionally t.ferred to tt some of the more aberrant forms. Cycauella wyomixgensis Ward. Three large specimens and many smaller ones in the last two nu-o.ees are referable either with certainty or with consideral.le probabthty to to peeies. One of the large ones. No. ,W0.88, consists o one large piece we gSng f, 8 kg. and two smaller ones of nearly equal size weighing togeth 1 1 9 kg.:all bearing the same number, and making a total weight of 8 7 k» These all fit perfectly together and form more than three- fourths 07 a very handsome trunk. The basal portion only wa.s wanUng, but the .reater part of it was found under two other numbers, vi , Nos. Cl .53 nd .5(10 525, which were fi«t found to fit each other and af ter- walo fill lost part of the large trunk. No. 5(K..513 weighs .72 kg ;■'. No. 5(,0..525 is a small piece weighing 177 grammes. Puttrng idl These pieces together we have a nearly complete trunk weighing H .(, kg. ThTo^ Iv part kicking is a small notch on one side of the base, which noire If the fragments in the collection would fit into. Thus comp^ e, th. trunk is very nearlv lh<. same size as the type specimen No. .,(«. o ..nd r^^ml es i,' in nianv respects. I. is, however, rather more Battened tog"t in the middle, where the major diameter is 24 cm. an.l the minor 13 em , giving a girth of .W .■m. I, stands 3ti cm. high. 184 :mi:soz()T(' floras of fxited states. \(). .")()(). (iSl represents a luuch larfjci' ti'unk, Imt is much less perfect than tlie one last described, and none of its missing parts could he found among the fragments. It consists of considerably more than half of the original trunk, which was symmetrically subconical, rounded at the summit, anil longitudinally flattened, but to a much less degree than the other. The principal loss is at tlie base by a rather even transverse fracture, but a large piece is also broken away from one of the thinner sides, the fracture deepening toward the summit and carrying away the apex altogether. It is now 24 cm. liigh and IS 1)V 25 cm. in diameter at the basal fracture, and has a girth of (H) cm. It weighs 11.34 kg. This, therefore, represents a larger trunk than any that were originally referred to this species. It also differs in some other respects from the types. The outer coating of ramentum is also wliolly removed and the surface thus exposed presents a somewhat different appearance from that of the other specimens. The l)ases of the petioles rise above the walls so as to give the surface a rough, warty aspect. The rock sul)stance is harder and heavier than that of the other trimks. These differences do not, however, seem to be specific, and I prefer to retain the specimen in this species. Xo. 100.227 is a segment from near the base of a still larger trunk. The basal fracture is nearly horizontal, Ijut the upper one is somewhat oblique, so that while the thicker edge is 12 cm. high the thinner edge is only 6 cm. This trunk was also somewhat elliptical, the diameters at the base being 19 cm. and 27 cm., respectively, giving a girth of 74 cm. The segment weighs 8.62 kg. The external surface of this specimen closely resembles that of No. 500.681 and it represents the same general type. These two specimens may ultimately require to be placed in a distinct species. PL XLVI is a side view, also showing part of the Ijase, of the trunk formed by joining Xos. 500.88, 500.513, and 500.525. A larger number of specimens had to be referred to this species than to any other, but all except those treated above are mere fragments, most of them quite small. Many of these fragments closely resemble the interior of the type specimens Xos. 500.7, 500.8, 500.20, and 500.67, and there can be no doubt that some of them are parts of the same trunk or trunks represented by those specimens, but in only one case has this been proved l)y finding the complementary parts. This case is that of s .(ihassk; cycads from Wyoming. isf) the small frasmont Xo. oOO.o'il, \vcij!;hin<:; 0.14 k lal(M-ac(|uire(l fi'auniciits pi-oved to he the complements of each other, viz. Xos. o()().17() and r)()().229. CyCA1)ELI..\ KNOWLroMANA WiU'll. PI. XI.VIl, Figs. 1, •_>. 1 !)()(). Cycadflla Knowltoniana Ward: Proc. Wash. Acad. Sci., Vol. I, p. 267, pi. xiv, figs. 1-3; pi. xviii-xx; Twentieth Ann. Rep. U. S. Geol. Suiv., ISOS-!)!), Pt. II, p. :?!)(), ])1. ixx. figs. 1-3: pi. xoi-xcv. This species, whicii, althoufih I'epresented by only two specimens in the original collection, furnished the l^est illustrations of the generic characters, is not absent from the material since received. The two specimens, Xos. 500.94 and 500.498, resemble the type X"^o. 500.76 sufficiently closely to have formed parts of the same trunk, Ijut this i; not proved by the discovery of any contiguous surfaces. Like that specimen, they are both somewhat triangular sections bounded by radial fractures, but showing considerable of the outer surface. In both, too, as in the tv'pe, it is the transverse fracture that l)est reveals the structure. If polished in the same w'ay they would both doubtless show all the characters of the genus. X"o. 500.94 is 8 cm. high and 13 cm. in diameter, which represents a chord of tlie cii'cumference, of which the arc is 19 cm., l)ut the surface is very irregular. The trunk was probably 15 cm. in diameter in this direction. The radial thickness is 10 cm., which seems to include more than half of the medulla, but this indicates a diameter in this direction of about 18 cm. The trunk was therefore elliptical. The fragment weighs 1.22 kg. X"o. 500.498 is a similar section, Init the fractures are all oblique to the axis. It is about 8 cm. high, 12 cm. in tangential direction, and the same in radial dii-ection. It weighs 1.41 kg. The outer coating of ramentum is nearly 2 cm. thick in places, and the armor 3 cm., the petioles and walls contrasting strongly in color, so as to show the structure to good advantage even on the unpolished surface of the fractures. Two other specimens, Nos. 500.102 and 500.285, are referred to this species, but these are one, since they perfectly fit each other, the latter being only a thin plate lying on one of the fractured planes of the former. 186 MESOZOIC FLORAS OF UNITED STATES. They thus form a re('tanfj;iilar piece from the middle of a trunk, showing the outer surface at both ends. The central part of the radial fracture is bounded by the inner wall of the woody zone and shows striations and markings not widely different from those much better shown in the type specimen. No. 500.62, which were described and figured. These specimens represent a trunk about 14 cm. in diameter. PI. XLVII, Figs. 1, 2, show, respectively, the outer surface of Nos. 500.94 and 500.498. Cy( ADELLA COMPRESSA Ward. PI. XLVII, Fifi. 4: PI. XLVIII. I'JOU. L'ycadeUa compressa Wartl: Proc. AVash. Acad. Sci., \o\. I, p. 269; Tweutieth Ann. Rep. U. S. Geol. Surv., 1898-99, Pt. II, p. 398, pi. xcvi; pi. xcvii. Five specimens in the second invoice and four in the third are referred to this species with more than the usual confidence, and a number with certainty. Nos. 100.228, 500.503, and 100.264 are nearly complete small trunks typical of the species. The first of these closely resembles the type No. 500.22 and the second the type No. 500.18. The third is smaller than any of the original types. No. 100.290 is a very small but apparently complete trunk weighing only O.IS kg., but it has the general character of this species. It may be immature. The other specimens are fragments, but No. 500.132 exactly resembles Nos. 500.68 and 500.69. Most of the rest have the same character. No. 500.503 weighs 0.74 kg.; No. 100.228, 0.75 kg., and No. 100.264, 0.31 kg. PI. XLVII, Fig. 4, and PI. XLVIII, Figs. 1, 2, show these three specimens, I'espectively, from their most characteristic sides. Cycadella jurassica Ward. PI. XLIX. 1900. Cycadella jurassica Ward: Proc. Wash. Acad. Sci., Vol.. I, p. 270; Twentieth Ann. Rep. V. S. Geol. Surv., 1898-99, Pt. II, p. .399, pl. xcviii-cxii. No large trunks of this species occur in the later collections, l)ut a very large number of fragments were foimd which can not be referi'ed elsewhere, and many of them show the typical characters so clearly that there is no serious doubt of their specific identity. .HKASSIC ('V('AI)S FROM WYOMINO. 1S7 The largest specimen is No. 100. 204. weighing 2.1 kg., hut this is supplemented by Xo. 500.507, weighing ().!)S kg., which nearly douhles the area of surface exposed and shows th(> charactcM-s more cleai'ly than any of the oi'iginal types. The specimen thus i-econstiiicted is a segment from the side of a large trunk, probably near tiic base, and extending <|uite to the middle. The diameter, which is still not complete, is 23 cm. Still more important was the discovery that the large fragment No. 500.511, weighing 1.54 kg., forms part of the type specimen Xo. 500.38, and exactly matches the fractured face represented in pi. cii of the first paper, completing the upper part. As a surplus of good foi- tune it was also found that the smaller fragment Xo. 500.516, weighing 0.65 kg., fits both these specimens in such a manner as to extend the part covered by its entire thickness of about 4 cm. This brings it within 5 cm. of the basal fracture, and we now have over two-thirds of the trimk. One small specimen, Xo. 100.250, weighing 0.43 kg., was found to belong to the type specimen Xo. 500.80, which was itself only a frag- ment and w'as not figured. It now becomes a respectable specimen. The other case of complementary parts is that of Xos. 100.289 and 100.292, which are both small specimens, weighing together only 0.23 kg. and showing nothing that is not better shown by other specimens. PI. XLIX is a view of the external surface shown iu the segment Nos. 100.204 and 500.507. Cycadella nodosa Ward. Pi. L: PI. LI; PI. LII, Fig. 1. I'.iOO. Cycadella nodosa Ward: Proc. Wash. Acad. 8ci., Vol. I, p. 271: Twentieth Ann. Rep. IT. S. Cieoi. Siirv., lS'.)S-99, Pt. II, j). 401, ])1. cxiii-exxii. Notwithstanding the small number of specimens in the later col- lections referable to this species and the fact that none of them supple- ment the original types, it has proved one of the most satisfactory of the specific groups. This is due to the fact that it contains two com- plete trunks that are quite as fine as any of the original types and two others that are very nearly complete. X^o. 500.509 is a complete trunk weighing 1.65 kg. and closely resem- bles the type Xo. 500.47, but is somewhat smaller. It has the con- 188 :\IKS()Z()T(" FTJIKAS OF rXITED STATES. tracted l):i.so oven more clearly marked. Xo. 100. 120(1 is also complete and weighs 2.83 kg. It closely resembles the type Xo. 500.17. It is nearly the same size as that specimen, but tlie rock is less compact and thei'e is some difference in the weight. Xo. 100.217 lacks the summit but shows the base very well. This was a larger trunk and still weighs 1.59 kg. Xo. 100.229 is a small, much flattened trunk, complete with the exception of a small piece. It weighs 0.47 kg. Except in size it is nearest to the type Xo. 500.21. The other two specimens are frag- ments and their reference to this species is somewhat doubtful. PL L is a side view of No. 500.509 and PI. LI a side view of X'o. 100.206. PI. LII, Fig. 1, shows one of the flattened sides of Xo. 100.229. Cycadella cirrata Ward. Pi, LII, Figs. 2-4; PI. LIII. 1000. Cycadella cirrafa Ward: Proc. Wash. Acad. Sci.. Vol. I, p. 272: Twentieth Ann. Kep. U. S. Geo]. Suiv., l,S9S-99, Pt. 11. p. 40:^ pi. exxiii-cxxix. A large number of small fragments in the second and third invoices show the peculiar internal structiu'e of this species, while there is nothing in their general character that negatives their reference to it. There are some others that seem to belong here, but which differ in the char- acter of the rock and in other respects from any of the types. If com- plete trunks of such had been found, it seems probable that they would have represented one or two new species, but lacking adequate material it seems best to refer them doubtfully to C. cirrata for the present. Many of these fragments are found to i-epresent complementary parts of one another, and a still larger number so closely resemble one another in structure that there is scarcely any doubt that they are from the same trunk that has disintegrated into small pieces. There seem to be three or fotu' such trunks, and nearly all the fragments can be referred to one or another of them. One of these trunks is undouljtedly the one to which the type specimens probably all belong (see the discus- sion following the description), but it has not been possible to find any exactly complementary parts of the types. The Xos. 500.136, 100.240, and 100.223 all join togethei' in that order, and Nos. 100.210 and 500.569 also join in such a manner as to make it practically certain that they all belong to one trunk and that JIKASSIC CYCADS FROM WYOMING. 18i> niilv n thin \nccv is iiiissintr between Xos. 500.136 and 100.210. If this were present, we should liave nearly half of a small compressed trunk. No. 500.470 joins Xo. 100. ;«S, and Xo. 500.504 joins Xo. 100.312. The-se la.st four, with about a dozen other frajjments, undoubtedly represent the type trunk, and ihe strui-ture indicates thai they lay veiy close to the type Xo. 500.71 (see PI. cxxviii). Xo.s. 500.178 and 500.422 constitute anothei' complementary couple of a somewhat different class, l)u1 evi- dently l)elono;inji to this species. They show the internal structure very clearly and also consitlerable of the surface, indicating a much compressed trunk probal)ly larger than that to which the type speci- mens belong. Xos. 100.258 and 100.275 also go together and represent still another trunk not othei'wise represented. They all have the char- acteristic internal structure and surface markings of the species. Xo. 100.245 is the lai'gest specimen, weighing 0.9 kg. It is a slab from one side of a trunk, showing considerable of the surface and a broad, smooth, tangential fracture. PI. LII, Fig. 2, is a side view of the triangular section i-esulting fi'om joining Xos. 500.178 and 500.422, and PI. LII, Fig. 3, the internal structure of the inner fracture of No. 500.422. PI. LII, Fig. 4, shows one of the fractures of Xo. 500. 130, and PI. LIII the tangential fracture of No. 100.245. Cycadell-v ExodEXA Ward. PI. LIV. I'JOU. ('!/cii(hllii exogcna Ward: I'nic Wash. Acad. Sci., Vol. I, p. 27.3; Twentieth Ann. Rep. V. S. Geol. Surv., 1898-99, Pt. II, p. 404, pi. exxx-cxxxvii. A limited munber of specimens, most of them fragments, are referred to this species. By far the largest and most complete is No. 500.514, which weighs 2.04 kg., although the armor has disappeared from more than half of it and a considerable part of the axis is also wanting. This specimen is difficult to orient, as neither base nor sum- mit is present and the direction of the axis is not clear, while the surface is covered with ramentum, but one of the fractures which passes through the entire woody zone and the armor was found to join one of the planes of fi'actui-e of the small specimen Xo. 100.249, adding considerable to the trunk. This .specimen weighs 0.45 kg., making the total weight 190 MESOZOIC FLORAS OF UNITED STATES. of the conihiiuition 2.49 kg. This specimen appears to have l)een nearly spherical, and somewhat resembles the trunk Xos. oUO. 1 9 and 500.53. Like that, the interior is very hard and smooth and the surface is covered witli a thick ramentum. Some of the fractures show the exogenous structure quite clearly. Xo. 100.220 is another good specimen, representing somewhat more tlian half of a small trunk and weighing 0.75 kg. The base and one of the sides are well shown and exogenous structure is clearly seen in the irregularly longitudinal fracture. Xos. 500.293 and 500.388 are two small complementary parts of the armor with regularly curA-ed iimer and outer surfaces, having exactly the same thickness and texture of the armor as Xo. 500.61, as shown in the transverse section, pi. cxxxii, fig. 2, of the first paper. They undoubtedly belong to that trunk and come from a point very near that fractured plane, but they do not exactly join it, and are probably from a somewhat different plane. They can belong nowhere else, as the opposite hemisphere is present in the specimen Xo. 500.53. These fragments are of unequal size and together weigh only 0.13 kg. There are two other specimens that complement each other, viz, Xos. 500.416 and 100.253. They are small fragments and weigh together 0.61 kg. They show the ring of woody wedges more distinctly than any other specimen except X^o. 500.19, l^ut they do not otherwise resem- ble that trunk and can not be referred to any of the triniks of this species thus far found. There is, however, no doubt that they represent this species. They may possibly belong to the same trunk as the fragment X*o. 500.174, which also shows the structure and the rings with unusual clearness. That specimen is a small segment from the base of a trunk weighing 0.44 kg. It extends to the medulla, and the fractures show three distinct rings of wood. Xone of the remaining specimens are specially worthy of comment. PI. LIV, Fig. 1, shows the tangential fracture of the first of these adjacent to the corresponding face of the second. PI. LIV, Fig. 2, shows the internal structure of No. 500.416, and Fig. 3 that of No. 500.174. JURASSIC CYCAD.S FKOM WYOMING. 191 Cycadella ramentosa Ward. PI. LV: PI. LXII, % 1. 1900. Ci/ca(Mla ramentosa Ward: Proc. Wash. Acad. Sci., Vol. I, p. 27.5, pi. xiv, fi-kno\vn species. It was represented in the original coUeclion of .Mr. W. II. !!(>(>([ hy only one specimen, Xo. 500.83, which is very anonialons in form and character, antl no one con d tell whethei" these wonld prove constant or not if more material were to come to li^ht. The pecnliar shape, resem- hlinsi a moccasin, was especially liable to vary. Neither was it pi-obable that the size wonld I'emain constant. It was interestinji, therefoi'e, to find this species rc^presented in the later collections. Among the fragments collected 1)>' Mr. (iilmore were fonr pieces, all of which bore so strikingly the characters of this species that I referred them to it before I discovered thai they belonged together. In the subsequent arrangement by species these were bionght into proximity, and I saw at once that they were parts of llie same tmnik. Two of the pieces had previously l)een found to fit each other and had been given the same number, No. 100.215. But one of the pieces also fits No. 100.202, and this in turn matches No. 100.230. The four pieces thus brought together form an almost complete trunk, which, though (x)nsiderably larger than the t^-pe specimen previously- known, has approximately the same anomalous shape, resembling a human foot. The compression, however, is more oblique, and the axis passes somewhat diagonally through the specimen. The greatest length is 28 cm., the greatest width 14 cm., and the maximum thickness 9 cm. It has a girth of about 35 cm. The four pieces together weigh 3.02 kg. There is a still larger specimen, No. 500.506, collected by myself, which I must either refer to this species or else treat as a new species. I pi-efer the former course, although it i-equires considerable explanation of the specific characters. It is nearly circular in outline and greatly com- pressed vertically, so as to have a somewhat lenticular shape with rather shai-p edges. The axis is slightly oblique ; at least it emerges considerably on one side of the center of the upper side. This axis is 12 cm. long, which represents the greatest thickness of the specimen. The diameter is about 2-1 cm. and the girth 75 cm. The surface is much obscured b\- com- 196 MESOZOIC FLORAS OF TNITED STATES. pression and distortion, hiil tlu' characters, so far as visiblo, are those of C. crepidaria. It weighs 6.13 kg. Xo. 100.203 is another practically complete specimen that must be referred to this species. It is subconical in shape, much flattened laterally, and tapers from base to summit. There are no fractures and the surface characters are well shown. Even the terminal bud is present and shows tlic summits of the small quadrangular leaf scars. The base is also per- fect, and here the axis is clearly separated from the armor all round. The two flat sides are unlike, showing that the trunk stood in an inclined position. It is 12 cm. high and the diameters of the elliptical base are 9 cm. and 16 cm., respectively, giving a girth of 39 cm. The armor varies in thickness from 5 mm. on one side to 5 cm. at the ends of the elliptical base. The axis also shows an elliptical cross section 5 cm. by 10 cm. in diameter. The trunk weighs 1.3 kg. No. 100.226 is a somewhat larger and less perfect trunk, but conforms more nearly to the type. It is flat like that, but the base is not well shown. Most of one of the flat sides is wanting and the mold of the medulla extends from the base to near the summit in the form of a hollow trough. The other side is perfectly preserved. It is 16 cm. high and had a major diameter of 14 cm. The specimen weighs 1.67 kg. Nos. 500.512 and 500.111, the latter quite small, exactly supplement each other to form another smaller l)ut complete trunk. It approaches the t_vpe more closely than either of the last two specimens described, in that the axis passes through the middle, transverse to the direction of greatest extension, but differs in the fact that the flattening is lateral instead of vertical when referred to the axis. Its \ery irregular form can be Best seen from the figures. Its height in the direction of the axis is 11 cm., and the diameters are I'espectively 9 cm. and 20 cm. The latter may be called the length. The girth is 45 cm. It weighs 1 .91 kg. The surface is black and rough, showing the scars imperfectly. The fracture between the two specimens shows that the interior is also black and the structure obscure. Five other small fragments have been referred to this species witli more or less confidence. PI. LVni is a \'iew of the liest side of the trunk consisting of Nos. 100.202, 100.215, and 100.230. PI. LIX shows the best preserved side of No. 100.203. PI. LX is a view of the convex side of the trunk Xo. 100.220. JURASSIC fVCADS FROM WYO.MkNG. I'JT ("ycadklla gki.id.v Ward. I'.IOO. Ciicadelhi fjd'ida Ward: Proc. Wasli. Ac-ad. .Sci., Vol. T, p. 2S1; Twontioth \\m. K<"p. r. S. Gool. Surv., ISDS-OO, Pt. 11, p. 414, pi. clxv cl.xix. From a resemblance in the leaf scars and the jj^eneral texture a number of small fragments found in tlie later collections are provisionally referred to this species. Cy< ADELLA CARBOXENSIS Ward. !!)()(). Ciicadella carbonensis Waril: Proc. Wa.sli. Acad. .Sci., Vol. 1. p. 2S;2: Twentieth Ann. Rep. U. S. Geol. Surv., lSflS-!)<), Pt. II, p. 41.j, i)l. clxx: pi. clxxi. In a few spots on the large type specimen, Xo. 500.2, the leaf scars ai-e seen, and their great size (3 cm. wide by 15 nmi. high) was one of the specific characters. Some are not more than 1 cm. high, with the maxi- mum width. One small fragment, No. 500.376, shows scars exactly like these. The texture and color of the rock are also the same, and it is tolerably safe to.refer it to this species. Cycadella Knigiitii Ward. PI. LXI. li)00. Cycadella Knigldli W'ard: Proc. Wash. Acad. Sci., Vol. I, p. 2,S.S, pi. xxi. Twentieth Ann. Rep. U. S. Geol. Surv., 1S9S-99, Pt. II, p. 41(1, j)!. clxxii-clxxvii. Only one specimen in the latter collection l)elongs with any certainty to this species, which was represented in the original collection by only two specimens, viz, Nos. 500.65 and 500.33, the former one of the largest and finest trunks found. The specimen now in question, though some- what smaller than that, is also a large and practically complete trunk. It is the Xo. 500.687, collected by myself. It is much compres.sed later- ally, deeply hollowed, and much distorted on one side. The other side is normal in shape and well preserved, but was so completely incrusted with lime that it was necessary to place it in a vat of muriatic acid for a considerable period. As this was the only side that .showed the scars sufficiently well for determination, I was uncertain until the lime was removed to what species it belonged. It proves to l)e C. Knightii, and is therefore the third specimen of that species known. It is consider- ably smaller than the type specimen, No. 500.65, and about the size 198 MESOZOIC FLORAS OF I'NITED STATES. of the less perfect specimen, No. 500.38. It i.'< 30 cm. high, 23 cm. in longer and 14 cm. in shorter diameter, and has a girth of 58 cm. It weighs 1 1.34 kg. PI. LXI is a view of tlie best sitle (that not distorted by pressure). INTERNAL STRUCTURE OF CYCADELLA. In the spring of 1901 Mr. (ieorge 1\. Wieland, after some corre- spondence with Professor Knight, in which the hitter authorized him to cut sections of the cycads from the Freezeout Hills, came to ^\'ashington and selected material for the purpose, which was sent to New Haven. Mr. Wieland has found time to make a somewhat careful study of some of the specimens, especialh" of certain ones belonging to the species Cycaddla ramantosa, the structure of which was also studied by Doctor Knowlton and myself. The results thus far obtained are important, and at m}' request Mr. Wieland has kindly furnished some notes describ- ing them and figures of the leaves detected in these fossils. I am verj^ glad of the opportunitj" to introduce his notes in full in this place and also his figures. ox THE FOLIAGE OF THE JURASSIC CYCADS OF THE GEXUS CYCADELLA. By Ci. K. Wieland. One of the most gratifying results of the structural stuch' of the fine series of silicified cycadean trunks from the Freezeout Hills of Carbon County, Wyo., constituting the genus Cycadella of Ward, has Ijeen the discovery of their young fronds. These have their structure preserv^ed, and also exhibit their prefoliation. Although minute and yet enveloped by the surrounding armor of leaf bases and ramentum, a fact to which we are chiefl}^ indel^ted for their preservation, the various tissues are already well developed, and one may sunnise with no small degree of confidence what must have been the character of the fully expanded adult frond. This is of importance because the occurrence in the fossil condition of foliage with structure preserved in connection with the trunks is rare. Of the vast numbers of isolated cycadaceous fronds with which the plant- bearing strata of Mesozoic age usually teem, wherevei- found upon the globe, only those of Williamsonia gigas (L. & H.) Carr. hiwv been foimd JrUASSIC ('V('Al)S FROM AVYOMING. lilO ill organic coiiiioctioii with the trunks to which Ihcv l)elong. And even in this case the evichniee u\Hm which WiUiamson based liis orifjinal rest ofat ion" was foi- many years calle(| in (juestion ])\ most working paleobotanists. Count Sohiis-Laui)acli (1887)'' states that the only instance known to him of a cycad witli attached knives that coukl he identified with certainty- was to be seen in a specimen of Willinfnsonid {Zamites) gigas from the Upper Jurassic sandstone of Yorkshii'e, jMiglaiuL This specimen was originally figured by Saporta.' Still further examples of more or less full-gi'own fronds of the \A'il- liamsonia type, whose organic connection with stems may lie j)roved, were given by Seward in 1897,'' together with (juite conclusive evidence in favor of the identity of Williamsonia and the cycadean trunks referred to the genus Bennettites of English and Continental paleobotanists. Here the subject rested until I announced the discovery of the crown of young leaves with structure preserved in the tjpe of Cycadeoidea ingens Ward.' This was one of the earliest results of the microscopic examination of our superb series of American fossil cycads undertaken by me, an examination which Professor Ward has done so much to encourage. Since then many additional facts have been discovered concerning the leaves of the Bennettitacea-, and the forms in various other species determined, but an account of these is reserved for a future extended publication on the entire sul)ject. It is only intended here to describe more briefly the discovery of the leaves in a typical form of Cycadella, this making the third Bennettitean genus in which they have been positively determined, and the second in which both structure and prefoliation are known — that is, if we regard Bennettites and Cycadeoidea as including forms generically distinct. Evidence is accumulating that such is the fact. « Coiitiil)iiti()ns towards the History of Zamia gigas Lindl. & Hutt., by W. C. Williamson. Trans. Linn. Soc. London, Vol. XXVI, London, 1870, pp. ()(i,'?-()74, pi. Hi, liii. b Einleitung in die Paliiopliytologic, Leipzig, 1887, p. 96; Introdnclidn to Fossil Botany, English trans- lation, 1891, p. 94. « Paleontologie fran(;aiso, Plantes Jurussi(|Ufs, Vol. II, Paris, 1873, p. 56, pi. lxx.\i, fig. 1. -30S, pi. vii. 200 MKSOZOIC FLORAS OF UNITED STATES. Leaves of cycadeiia ramentosa Ward. — A closcr examination of the middle one of the three .so<;nients figured by Professor \\'ard in the first paper on the Status of the Mesozoic floras of the United States" possibly representing the main portion of a typical specimen of this species, resulted in the discovery of the two fronds shown in transverse section in PL LXII, Fig. 2, and PI. LXIII, Fig. 1. As is represented in the figures, looth of these fronds are very small and are entirely surroimded by remarkal:)ly preserved ramentum. Both these leaves, as will be seen on examining Professor Ward's figure, show'ing very neatly the middle (not terminal) position of the portion of the trunk which bears them, are abnormal in being borne laterally, and not as members of a series forming a crowai of leaves. They grew out, therefore, from between old leaf bases, probably after the crowai of the plant had suffered some injury. As this phenomenon has been observed in some other cases not yet described, I regard it as possible that these plants may have been subject, among other mishaps, to cropping by contemporaneous animals (dinosaurs?). However, this abnormal position does not affect the orientation of these leaves. They are normal in all other respects. Their prefoliation agrees with that of Cycadeoidea ingens, as descri))ed by me. As in that species, they are once pinnate, the petiole being distal and the pinnules'" folded back face to face in tw^o ranks. But in structure there are certain points of differ- ence. In Cycadeoidea ingens the hypodermal sclerenchyma beneath the upper surface of the leaf is continuous with the bundle sheath. In the present specimen there is no such connection, although the bimdle sheath of sclerenchyma is strongly marked, as in the living cycad Ccrnioznmia fuscoviridis [rz C. Mexicana Brongn.]. The prefoliation and arrangement of parts in the present fossil form bear an exceedingly close resemblance to this living form, the closest I know. But, on the other hand, Ceratozamia terrestris, with a strong development of hypodermal ribs of sclerenchyma on both the xylem and phloem side of the pinnule bundles is rather more like Cordaites ("?) anguloso-striatus Gr. Eury., as figured by aTwentieth Ann. Rep. U. S. Geol. Surv., Ft. II, 1900, pi. cxxxviii, fig. 2. I> These leaves are once pinnate, but since wc liavc bipinnuto cycad leaves (Bowcnia), 1 sliall use the term pinnule rather than pinna. This is the usage in some of our most recent and best pubhcations on ferns, and the terminology of both must obviously be homological. Moreover, this is the more convenient usage, since in the case of many fossil forms the order of the lamina; is not always readily determinable. .IIHASMC CYCAPS FIJOM WYUMIXCJ. 2Ul Renault (('our. I5..1. I'oss., 1, pi. xii, fig. 3)." tliaii like eWhev Cycadeouica ingcn.^ or Cycnddla nimautoxa. That is to say, the differeiu-es l)etween tliese \-arious forms are not such as in themselv(^s ina>- l)e very readily used for othei- than specific separations. Tli(> fact is also emphasized that we are hei-e tlealing with a very generalized type of leaf structure. Bundle pattern of the petiole.— This IS IndlCatcd \\\ t llC VOUUg IcaVCS of C IJCd- (Iclhi ramcnto.^i and differs markedly from that of the living cycads. It is nuich more fern-like in l;)eing arranged like a V with a ([uite continuous xylem and phloem region. It is desired to cut some further sections l)efore giving detailed figures. In PI. LXIII, I'lg. 1, there is probably shown some distortion l)oth of tlie l)undle region and of the transverse section of the petiole, due perhaps to desiccation before the process of silicification began. The leaf represented in Fig. 1 of the same plate shows, in the uniform contraction between the veins of the pinnules, what is quite likely an abnormality due to the same cause. The bundle pattern of the petiole is not to be confused with that of the leaf base. The latter, of course, merges into the former. Number of pinnules.— This has uot beeu determined, since it has not l)een thought desirable to sacrifice any of these leaves by cutting them in longitudinal section, especially since lumiber is scarcely of specific value, varying often in the case of fronds from the same plant and very greatly as a plant grows older and the relative size of its fronds increases. The numloer in the present species may have been as low as 30 or doubtless as high as 60. Form of the full-grown pinnule. — Thls uiay bc surmlsed with no small degree of confidence. The fact that in the case of the frond shown in PI. LXIII, Fig. 1, the number of bundles increases as successive pinnules are cut, and then becomes constant, proves that the venation is dichotomous after the manner of Zamia angustifolia Jacq., and shows that the pinnules, though elongate, can not have had a broad base with numerous veins like Dion, nor yet Hke Williamsonia gigas. (See PI. LXII, Figs. 1 and 4.) The degree to which the tip may have been acuminate and the relative width are of course not possible to deter- mine from transverse sections alone. But undoubtedly the pinnules of the present species presented an appearance quite like that of such a This figure is reproduced in Scott's Studies in Fossil Botany, p. 423, fig. 139A. 2L)-2 MESOZOR' floras of I'MTED STATES. forms as Podozamites lanceolnius (L. & H.) Fr. Br. fi-oin the Jurassic of Oroville, California; P. laticeolatus lafifolms (Fr. J^r.) Iloor, or P. Emmonsii Xewb., from the Trias of Xortli Carolina." PI. LXII, Figs. 1-3, PI. LXIII, Fig. 1, Cijcadella ramentimi Ward; PI. LXII, Fig. 4, Williamsonia gigas (L. & H.) Carr. PI. LXII. Fig. 1. Hypothetical form of portion of mature frond, based on transxerse sections shown in Figs. 2 and 4. From the sec- tions of the yoimg fronds we learn that the frond was once pinnate and the bundle system strongly dichotomous. Further, while the exact form is somewhat conjectural the successive increase or decrease in the width of the several pinnules, as cut transversely, pei-mits a nearly correct interpretation. (See Fig. 5, showing the best known related form.) PI. LXII, Fig. 2. Transverse section of a very young frond embedded in ramentum, only partly shown. The position of the petiole is shown in dotted line. The closely folded pinnules show a series of ridges cor- responding to the venation and bundle system, the bimdles being indi- cated in the drawing l)y small circles. The ridges are probably due to some condition attendant upon silicification. X 25. (See PI. LXIII, Fig. 1.) PL LXII, Fig. 3. Camera lucida drawing of transverse section of the ramental chaff or flattened hairs enveloping the still folded young fronds shown in PL LXII, Fig. 2, and PL LXIII, Fig. 1. These hairs were several inches in length and a single cell in thickness at their origin. Well out toward their tips they are three and four cells in thickness, as here shown. X 65. PL LXIII, Fig. 1. Transverse section of an emerging young frond 1 cm. distant from that shown in PL LXII, Fig. 2, but larger and lietter preserved. The pinnules with their bundles indicated are folded l:)ack to face, in two ranks, this indicating a once pinnate frond with the pre- foliation of Cycadeoidea (see Wieland, loc. cit.) and the living Dion. The somewhat furrowed (or dried?) rachis is seen at the lower right- hand corner, the interior stippled area marking the fern-like bundle system. The arrow points toward the central axis of the trimk, the rachis being distal. X 25. « Professor Fontaine's figures of these forms maybe consulted. They are respectively given on pl.l.xiv, figs. 1 and 2, and pi. xlii, fig. 1, of the first paper on the Mesozoic Floras of the United States, Twentietli Ann. Rep. U. S. Geol. Survey, Pt. II, 1900. .11 KASSIC CVCADS FliO.M TllK I'.LACK IIILl.S. 203 PI. L.XIl, I'itr. I. Scnvard's figure of ;i fine fi'oud of W illidmsonia gi'gn>< ( I.. iV Il.i ('an-., pi. \- of the .Iiii-assic I'lora of ^'ol•kshil•<^ Part T. Natural size, cf. PI. LXII, V\ii. 1. Note.— PI. LXII, Fio-s. l-;i, ami PI. LXllI. 1' i,u-. I. aic from the Univoi'sity of \\'yomini' Ci'etaceous (Lakota formation of Darton). This last is the soui'ce of the great numbers of eycadean tmniks that 1 have deso-ihed from the Blaek Hills. These occui- about midway of that formation, and below the cycad horizon are \-arious plant beds containing impressions of c.ycadaceous vegeta- tion. Until recently no plants except fossil wood had been found in the underlying Jurassic t)eds, the upper member of which is the Beulah formation (Beulah clays of Jenney), in which occur the Atlantosaurus beds of Marsh. When I made my fourth and last visit to the Yale Museum, in May, 1900, to complete the elaboration of the great collec- tions of cycads that Professor Marsh had so mvmificently accumulated there, I found one ver}' anomalous specimen that had been purchased for Professor Marsh l)y Mr. H. F. Wells from a dealer in Hot Springs who had ol)tained it from a stranger and had no record fiu'ther than that the man who sold it to him had told him that he obtained it "50 miles west of Hot Springs in Wyoming." I named the new species, which it clearly constituted, Cycadeoidea utojdensis, but in the descrip- tion 1 stated that on the surface there was "an area near the summit covered by what appears to be an outer coating of ramentum, as in the genus Cycadella, more or less obscuring the organs." At the end of the discussion I .said: "The patch of I'amentuni, if such it be, near the summit of the specimen, raises the suspicion that it may l^elong to the genus Cycadella, and, as all the specimens of that genus thus far known have come from the Jiu'assic, it is possible that the horizon of the bed holding this specimen may be lower than that of the other Black Hills cycads." I also discussed the piobablc locality and regarded it as "more pi'obable that the direction was northwest from Hot Springs, 204 MESOZOIC FLORAS OF UNITED STATES. and this might locate it in the Lakota formation some distance north of Cambria and in the general region of the Newcastle coal field."" Mr. George R. Wieland, who has taken a deep interest in all matters relating to fossil cycads, whose internal structure he is so successfully working out, spent a good part of the field season of 1900 in the Black Hills making collections for the American Museum of Natural History. He paid special attention to questions of stratigraphy, and made many valuable sections, which, through the kindness of Prof. H. F. Osborn, I have the permission to use in this paper, together with other informa- tion which Mr. Wieland, at my request, has contributed. He studied the Jurassic and Lower Cretaceous beds of the Black Hills on nearly all sides, but especially on the northeast and southwest sides. In the course of his investigation of the Jurassic beds northwest of Cambria he discovered fragments of cycads in the Beulah clays, occupying a stratigraphical position similar to or identical with that of the cycad bed of the Freezeout Hills in Carbon County. This locality is between 50 and 60 miles northwest of Hot Springs, and therefore corresponds, in distance, at least, to the source of tlie Cycadeoidea utopiensis. He sa}-s that the specimens obtained there by him resemble that speci- men. He has also carefully examined the patch of ramentum on that specimen described by me and has no douljt that it belongs to the genus Cj^cadella. There is scarcely anj^ douljt that all this is true, that he has virtually found the locality, and that the specimen reall.y came from Jurassic beds. The species is therefore transferred to that genus and will henceforth bear the name Cycadella utopiensis (Ward) Wieland. It is figured in this paper on PI. LXIII, Fig. 2. Mr. Wieland has furnished the following notes and sections relating to the geolog>' and paleontology of the southwest side of the Black Hills in Crook County, Wyo., which are of special interest in this connection : " Elaboration of the fossil cycads in the Yale Museum: Am. Jour. Sci., 4th ser., Vol. X, November, 1900, pp. 3"27-.34.5, pis. ii-iv. Cycadeoidea utopiensis is described on pp. 338-340 and figured on pi. iii, upper figure (No. 727 of the Yale MuseumJ. .irPvASSIC CYCADS FROM THE I'.LACK HILLS. 205 FIELD NOTES. By (i. I\. W'iKLAMi. A comparison of the hods on the soutlnvest side of the Black Hills willi those on the east and northeast sides shows that in the former th(» beds lie nearly horizontal and are deeply cut by streams, so that llie lineal' exposure of tlie Jurassic is immensely increased, lliv most marked chanjje is in the diminution of the sandstone l)ed beneath the main Atlantosaurus beds, or Beulali shale, if, indeed, the 25 feet of yellowish sandstone here intervening between this bed and the marine Jurassic may be considered equivalent to the ''Unkpapa" of the east- ern hills. Antl, conversely, there is an increased thickness of the over- lying l^eulah shales. Thickest on the northwestern side of the hills, and absent from nearly opposite Hermosa to Minnekahta, this bed, teeming at its base with the remains of huge dinosaurs, incloses the Black Hills like a long-armed crescent or horseshoe. Most unfortunately these saurian bones are seldom well preserved and the collector is always baffled by one long stretch of talus after another. It will prove possible, however, in the course of time, to deter- mine the extensive fauna represented, and in part its silicified fiora of cycads and conifers. This being, of course, an easterly extension of the Jurassic so well marked farther west, most of the forms are doubt- less already known. The section is of especial interest in connection with the stratigraphic relations of Cycadeoidea and Cycadella. Section .} milen irent of Ilulett, Crnoh Count)/, W;/o. Feet 7. Various clays or sIihIps hiuI siindstoiics containing some silicified wood, iiiul doiiljl less in tlieir lower por- tions the equivalent of the Blackhawk and Minnekalita cvcad beds (overlain unconforniahlv liv the Fort Benton Cretaceous) 2IXJ 6. Black sliales eoiitaiuing more or less distinct remains of dinosaurs 30 3. Bluish shale weathering white. Contains remains of large dinosaurs, .seldom well preserved, silicified wood, and probably cycads 12 4. Yellowi.sh to red shale s ',i. Clay containing three or four thin nodular layers with remains of largo dinosaurs, and ending rather sharply below as light sandy or nodular material 40 2. Sharply defined stratum of yellowish sandrock, barren, so far as observed 2o 1. Marine Jurassic, ending above in limestone weathering whitish and containing remains of Baptanoilon (and Megalosaurus ? ) 200 Total (approxiuuite ) ,5,3, o 200 ]\IES0ZOl(^ FLORAS OF CXITKn STATES. This is a nnich less complex section than obtains southward from Hulett in the direction of Newcastle, in which direction the Bevilah shale series especially is more highly developed than elsewhere about the hills. But it is representative. That somewhere near bed No. 5 in the Beulah shales cycads are found is proved by several specimens which occui' in connection with numerous remains of large saurians on the Anderson ranch, near the head of Skull Creek, 4 miles south of Inyankara Mountain. Botli the fossil Ijones and the cycads, as well as much silicified wood, plainly l)elong near the base of the ];)est marked shale seen at this point. These cycads belong to the genus Cycadella, and are the first to be definitely located in the lower fresh-water Jurassic of the Black Hills. The fine trunk No. 727 of the Yale collection, named by Professor Ward Cycadella utopiensis, and originally labeled as having come from "50 miles west of Hot Springs," a very unlikely locality, doubtless came from the In^-ankara Mountain country. These specimens have precisely the type of preservation seen in the Cycadellas from the Freeze- out Hills of Carbon County, Wyo. Both weather white and fracture black, with the same characteristic surface and shades, as do also the segments of silicified tree trunks not only common to both these cvcad localities, but plentiful also on the eastern side of the hills. Not alone, therefore, from the general character of the Beulah shales of the east- ern hills, but from the testimony of the most striking forms of animals and plants as well, must we consider them the easterly extension of the cycad horizon of the Freezeout Hills. A few feet over this horizon I believe Professor Marsh's Jurassic mammal quarries to have been located, and the cycads of the Blackhawk and Minnekahta localities in South Dakota must occur from 75 to 125 feet higher. The preser- vation and character of Professor Ward's Cycadcoiclea heliochorea from 9 miles northwest of Sundance prove that it is a species belonging in the group of cycads obtained in the Minnekahta region, and that it has doubtless come from the same general position, though the specimens thus far obtained, like the great majority of the cycad trunks, were not found actually in place. The free development of ramentum and the uniformly small size of the older or Cycadella series, as compared with the younger and larger NOTK ON FIJKMONTS COI.LPXTION. 207 Cycadeoideas, is a poiiil llial al once su.ii.u'cstscithci' a cliaiifje in cliiiialcMii the inlci'N'aJ IhMwccmi ihc tossilizal ion of the Iwo li'roups oi', if they con- linucd 1() exist near (\'i('li oIIkm' in time, tlic pi'escncc about the .lui'assic frcsii-watei' lake of areas with diverse cliinatie conditions. In ttic Black Hills count I'v, at least so fai' as I have noticed, thei'e is likewise a coi're- spoiidinfi difference in the size of the acconipanyinij; silicified tree trunks. Many of the .Vraucai'ioxylons of the upper c\'cad IxmI wore of iininense heiiiht and size. I'ivalintr the Norfolk Island pines of the jiresent. while, on the contrary. 1 have never found markedly larjie tinniks in coiuiection with th(> ( ycadella horizon, although they may occur farthei' west in Wyoming. That any great change took place is not argued. The Cyca- dellas may have grown in dry or arid situations, oi' in a climate like that of Floi'ida, where the dwarf Zamias thrive wliile the Cycadeoideas l)espeak moister and more distinctly tropical conditions. I first pointed out, in a review in the .\merican Journal of Science for May, 1900, j^age 386, the fact that Cycadella might have grown under le.ss favorable conditions. Seward likewise considei's that tJiey may afford suggestions of value con- cerning climatic conditions." The microscopic study of the Cycadella series which I have under- taken in conjunction with the study of the Cycadeoidea has not been completed, so that later it will perhaps be possible to discuss such ques- tions as this with more safety. In connection with the relative position of Cycadeoidea and (Vca- della, 1 should here mention that, under the direction of Professor Osl)orn, Dr. F. B. Loomis has prepared especially satisfactory sections of the better exposed Jurassic beds as seen farther west in Wyoming.'' While this paper gives more exact information concerning the successive hori- zons and their vertebrate fossils, no mention is made of plant remains. I conclude that the bed Doctor Loomis numbers 22 is the cycad horizon of the Freezeout Hills. XOTE ox FKEMONT'S COT.T.EOTION. In concluding this account of the known Jurassic floras of the United States, it is well to call attention to the collection made by Fremont on "See Nature, October 24, liX)!, p. («:! ''Oil Jurassic stratigraphy iu soullicastern Wyomiiii^: Bull. Am. Mus. \ut. Jli.st,, \ol. Xl\, Article XII, pp. 189-197. 208 MESOZOIC FLORAS OF UNITED STATES. August 19, 1843, which was elaborated Ijy James Hall and the age pro- nounced Oolite by him. Part of this collection is in the National Museum, including the principal types figured by Hall, and I long ago sent it to Professor Lesquereux, who examined it and gave his reasons for agreeing with Hall, notwithstanding Heer's opinion to the contrary. Hall admitted the existence of one dicotyledonous leaf, which should have sufficed to prove his error as to age. A single glance at the leaves called Glossopteris Phillipsii by Hall is sufficient to show that they are dicotyledons. I have already given references to all the papers dealing with these plants," V)ut should have referred to the page of Fremont's report on which the itinerary- occurs, viz, page 131. The locality is near Evanston, Wyo., and the plants probably occurred in the Bear River beds, which are certainly Cretaceous and even Upper Cretaceous, but not Laramie, as was once sup- posed. Dr. C. A. White and Dr. T. W. Stanton place the Bear River beds between the Dakota and Colorado formations.'' a Eighth Ann. Rep. U. S. Geol. Survey, Pt. II, 1889, p. 870. b On the Bear River formation, a series of strata hitherto known as Bear River Laramie, by Charles A. White: Am. Jour. Sci., 3d ser., Vol. XLIII, February, 1892, pp. 91-97. The stratigraphic position of the Bear River formation, by T. W. Stanton: Op. cit., pp. 98-115. F^J^T IIT. THE CUIETACE0U8 FLoKA. Tn colli iiiiKit ion of the pl:ui of this series as outlined in the introdue- toi'\- ivinai'ks to the first paper, the treatment of the Triassic flora (Part 1) and the Jurassic flora (Part II) having l)een completed and all available infoi-niation with i-egard to them having been brought down to date (close of the year 1903), the Cretaceous flora (Part III) may now be taken up. In eny Sir William Dawson, including the following species from the (^ueen Charlotte Islands: Osmundites skidegatensLs Penh. Neuropteris heterophylla Brongn. Ta?niopteris plumosa Dn. Sagenopteris Nilsoniana (Brongn.) Ward. Sageiiopteris oblongifolia Penh. n. sp. Sagenopteris elliptica F'ont. Zaniites crassinervis Font. Zaniites tenuinervis Font. Nilsonia polymorpha cretacea Penh. u. var. Ginkgo pusilla Dn. Sef|uoia Langsdorlii (Brongn.) Heer. His only figures are of internal structure, .vhich does not usually give specific characters, and only three of the species are even thus illustrated. Some of the names are prima facie doubtful, e. g., Neuropteris heterophyUa and Sagenopteris Nilsoniana, the first a Carboniferous species, and the other Older Mesozoic. These at least should be figured, that one may judge l)etter of the age of the formation. His Nilsonia polyniorplia cretacea, which he calls a new coml)ination, but which seems to l)e a new variety of his own, is also doubtful. He cites the figure in Schimper's Atlas, pi. xlv, fig. 6 (copied from Schenk's Flora d. Grenzschichten, pi. xxix, fig. 11), from the Rhetic of Franconia. If he has such a leaf it is strong evidence of at least Jurassic age. a Geological Survey of Canada, Report of Progress for 1878-79, Moiilri'nl, 1880, pp. 1-23'JB. ^ Trans. Roy. Soc. Canada, Soct. IV, Vol. VIII, pp. 3-29, pi. i-vi (=pp. 19-29). c Pages 31-91, pi. vii-.wi (=pp. 73-91 ). THE CRETACEOUS FLORA. 211 FI.OItA OF THE ISIIASTA KOUMATIOX. Fossil plants liave l)OCii found in tlu> Shasta bods in both California and Oregon. I'ntil refontly there was ijreat confusion in the plant- bearing beds of Oregon, as it was not supposed that the Jurassic was found there. As shown in this paper, however, all the specimens fi'oni the Buck Mountain region, as well as those from the Cow Creek Valley, ncai- Nichols station, came from the Jurassic. Those, howevei-, from localities farther east, especially from neai' the town of Pviddles, ai-e of Shasta age and will be treated mider this head. During the progress of the topographic survey of the Red Bluff (juadrangle, in Shasta and Tehama counties, Cal., in charge of Mr. Gilbert Thompson, which was made in the years 1882-1884, Mr. Thompson found a plant-ljearing bed near Pettyjohn's ranch, on the Cold P'ork of Cotton- wood ('reek, Tehama Coimty, and collected and sent in a numl)er of specimens. Only one of these, however, seems to have been saved, and this was sent to Prof. Leo Lesquereux, who determined it as a Pecopteris, without assigning to it a specific name. As such it was duly recorded in the catalogue of the National Museum as No. 2193. It was in two parts, completing each other, and these have been glued together. These parts beai' Professor Lesquereux's numbers 254 and 255. Owing to the obscure chirography of the label, the name of the locality was misspelled in the Catalogue and the attention of the geologists who subsequently studied the beds of t his region was not. called to it. As soon as the correct name, Pettyjohn's ranch, was known, the specimen, which had long lain in a drawer waiting for data to fix its position in the collections, assumed a special interest and steps were taken to learn more of its historw It was shown to Mr. Gilbert Thompson, who recognized it at once and distinctly remembered collecting it. He indicated the epcact locality- on the map, which would certainly place it in the Shasta formation and well up in the Horse town beds near the base of the Chico. The character of the rock agrees well with this and there is nothing remarkable except the fact that the plant seems to represent the chiefly Paleozoic genus Pecopteris. It is a large, distinct fern, wholly unlike any of the others that were collected in that region. It may well have been a tree fern. As Professor Fontaine says, the finer nervation is not shown, and it is still possible that it may belong to some of the Mesozoic genera to whicli '■) I •) MESOZOIC FLORAS OF UNITED STATES. in;iiiy of the forns of that age, first regarded as l)elongiiig to Pecopteris, have recently been referred. In the spring of 1SS7 Dr. (\ A. White showed me a specimen from the Shasta formation of Ctdifornia that had come into his possession without any more definite indication of its exact source. I sent the specimen, on April 4, to Professor Fontaine, who re})lied: "The specimen * * * is a Sagenopteris. I can not distinguish it from Sagcnoi)icn'fi eUiptica sp. nov., the most abundant sj^ecies foimd in the Lower Potomac group of Virginia." This species, as will be seen, occvu'red in later collections, but is not very common. The next earliest record we have of the discovery of fossil plants in the Shasta formation is that of a few specimens turned over to the division of paleobotany' of the United States Geological Survey bj' Dr. T. W. Stanton on March 17, 1890. They appear to have been collected the previous season by Mr. Will Q. Brown, and were found in the Hoi'se- town beds, in the vicinity of Riddles, Oreg. Two other specimens were received in February, 1892, from Mr. J. S. Diller, collected alscj by Mr. Brown, in 1891, from the same locality, viz, "on Cow Creek, close to the town of Riddles." Mr. Diller sent two other specimens direct from the field in Jtme, 1892, also from near Riddles. In 1893 Doctor Stanton and Mr. Diller, assisted by Mr. James Storrs, made extensive collections from the Knoxville and Horsetown beds of California on the eastern slopes of the Coast Range, drainage of the Sacramento River, below the latitude of Mount Shasta. They fovnid an abundant fauna, but the flora was meager. Still, their collections of fossil plants were rather large and came into my hands before the end of that year. I made a preliminary report upon them, jjut was obliged, for want of time, to send them to Prof. Wm. M. Fontaine for more thorough examination. He reported upon them somewhat fully untler date cf February 23, 1894, and his identifications were published !)>' Diller and Stanton in their paper read before the Geological Society of America, which, though read on December 27, 1893, or before the report was com- pleted, was not published until April 12, 1894." " Tlie Slmsta-Chico series, by J. S. Diller and T. W. Stanton : Bull. Geol. Soc. America, Vol. V, Rochester, April, 1894, pp. 435-464. See pp. 4.50, 451. ' FLORA OF TlIF SHASTA FOKMAl'lON. 213 III ISO! Doctor Sliiiitoii collectod two simm'Uikmis in tho vicinity of Ivuldlcs, on tiic left l)ank of Cow Crook, which was practi(;ally the same locaUty as that of most of Mr. Brown's coUoctions, th()ii, who repoi'ted on tluMU to Doctor Stanton under date of March 12, LSOo, and tlie idei:- tifications w(>re published soon aftei'." Tlie fragmentary character of the material from all these l)eds, wliich rendered most of the determinations more or less doubtful, made il desirable to have a more special search made foi' vegetable remains, and at the urgent request of Mr. Diller I decided to spend some time in ('alifornia with this oljject in view. I secured as accurate information respecting the localities as possible and joined Mr. Diller's })ai1y at Ivoseliurg, Oreg., on September 0, 1895, after having made the coll(>ction of Kootanie plants at Great Falls, Mont., to be described later. It. was arranged that Mr. Storrs should accompany me to the localities in California. I did not stop at Riddles, as the importance of the fossil flora of that region had not yet been emphasized, but proceeded to Ono, Shasta County, Cal., where Mr. Storrs soon joined me, and we spent sixteen days in the general region where fossil plants had been previously obtained. Ono was made the base of operations from September 9 to September 15, and the principal localities in that vicinity w'ere very carefully exam- ined. Fossil plants were found in Byron Gulch, northwest of Ono and close to the town; on Cottonwood Creek, below the mouth of Eae:le Creek and above that of Hulen Creek ; southeast of Ono ; and in Aldersons Gulch, 2 miles soutlnvest of Ono. We also spent a day in the vicinity of Horsetown, wdiich is 8 miles east-northeast from Ono. Two miles northeast of Horsetown, on the road to Centerville, at the southern base of a nearly east-west ridge, the Cretaceous is exposed, overlain ])>^ a mass of tufa. In the fine-grained concretionary rocks that occur among the shales, much as they do in the region around Ono, we made a fairly good collection of plants, mostly conifers. The localities near Ono yielded comparatively little. A few fern fragments were found in Byron Gulch. Coal was reported on Cotton- « Contributions to the Cretaceous paleontology of the Pacific coast ; the fauna of the Knoxvillc lieds, hy T. W. Stanton: Bull. U. S. Gcol. Survey No. 133, 1895 (issued February 3, 1896). See p. 22. 214 MESOZOIC FLORAS OF UNITED STATES. wood Creek, and a man who had seen it guided us to the locality, which is on Cottonwood (^reek, a quarter of a mile below the mouth of Eagle Creek. As I expected, the (!oal proved to be lignite, and there are many lignitized as well as silicified logs and quantities of blocks of fossil wood. The trunks are frequently silicified in the center and lignitized near the surface. The \\ood usually shows the grain well. Many of the rocks in which the trunks are embedded contain vegetable matter, mostly in the form of coaly stems. A few recognizable plant impressions were, how- ever, found, chiefly fragments of ferns and leaf-l^earing coniferous twigs. As the strike of these beds is here northeast-southwest and the dip to the southeast is very steep even here, though much less so than farther south, the strata rise rapidly in descending Cottonwood Creek, and there is a correspondingly rapid change in the character of the flora. The strata could not be traced continuously, but at the mouth of Hulen Creek, 100 yards above the junction of the two streams, in coarse, dark- colored sandstone shales, dicotyledonous leaves were found. Owing to the coarse matrix, the nervation is obscure and the material obtained is •rather poor. This bed belongs to the Chico, according to Doctor Stan- ton's deterininations, and these dicotyledonous leaves are not included in the descriptions given in this paper, l)ut the material is reserved for a later paper which will treat exclusively of the upper leaf-bearing beds of the Lower Cretaceous. In Aldersons Gulch plants wei'e exceedingly scarce, but in two places we found them in the hard, fine-grained concretionaiy rocks that ever>Tvhere foiTn seams among the shales. Several coniferous twigs were found, a few showing the leaves. Cycadaceous vegetation was also detected. Fossil wood is abundant. On the 16th we left Ono and proceeded southward to Stephenson's ranch, on the Cold Fork of Cottonwood Creek, a mile above Pettyjohn's, in Tehama County. A few fossil plants were found on this stream at two localities above the ranch, chiefly ferns and conifers. Vegetable remains are here very rare. From here we continued om- journey southward and arrived on the 18th at Lowry, on Elder Creek, which was made the base of operations during the remainder of the expedition. Four miles west of Lowry, on the North Fork of Elder Creek, plant remains were found at several FLOKA OF TIIK SHASTA FORMATION'. 215 localities and horizons, hut usually in \cvy small f ragmen ts, consisting of the tips of the pinnules of ferns, cycads, ete. On the South I'ork of Elder Creek, from li to 2 miles above Lowry, at and below tlu^ dam, plants also occur in nuich the same condition as at the last-mentioned localities. Fai'ther up the South P'ork, l)elow Coopers, 5 miles south- west of ]>owrys, near the gorge whei-e IIh> South fork cuts tlu'ough a heavy bed of conglomerate, we found a t)ed that yielded fei-ns, cfcadaceous leaves, etc., and made a considerable collection. The most southern point visited was Wilcox's ranch, (5 miles soutli of Lowry, and over the divide between I'Hder Creek and Thome ('reek, on the road to Paskenta. Mr. Storrs had found one specimen here on a fomier occasion, l)ut we were unable to find any more at the original locality. At another place, half a mile east of Wilcox's, we found some very imperfect fragments. Just on the crest of the divide, about midway between Lowry and Wilcox's, a bed was discovered l)v the roadside containing delic^ate fern impressions and detached cycadaceous leaflets. They occiu' in a rather fine sandstone shale, slightly concretionary, and were found on both sides of the road, but chiefly on the east side, at two horizons 80 feet apart. A large number of specimens were obtained. A single dicotyledonous leaf had been collected by Doctor Stanton in 1893 from a locality 2^ miles below, or to the east of Lowry, on Eldei' Creek, and Mr. Storrs and 1 tried to find the spot, but prol^ably failed. At least we found no fossil plants in that general region. Half a mile above, however, on the left bank of Elder Creek, a few fragments were broken out of a coarse sandstone ledge, one of which was a small fern, and the rest seemed to be pine needles. A much better locality, and one that had not been previously discovered, was at the eroded end of a low ridge run-ning north from Elder Creek, only half a mile below Lowry. In a light-brown sandstone ledgie at this point there occur well-preserved dicotyledonous leaves and some other vegetable impressions, of which we made a considerable collection. As in the case of the dicotyledons found at the mouth of Hulen Creek, these leaves are reserved for a future paper. The last-named locality belongs to the upper Horsetown beds according to the sections that have been made, but the specimen collected by Doctor Stanton 2 miles below comes well up in the Chico. It was on 216 MESOZOir FLORAS OF rXlTHD STATKS. Elder Crook that Mr. Diller made his woll-knowii section," which so staggered the geologists who are studying the age of the earth. Accoi'd- ing to this section the Knoxville l)eds haAT a tliickness on Elder Creek of 20,()(30 feet, the Horsetown l)eds of aljout (),000 feet, and the Chico beds of 4,000 feet, making 30,000 feet measurcHl. And yet it seems that "the complete series of the Shasta-Chico beds is not exposed in this section.'"' »The strata here often approach a vertical position and the section stretches across their upturned edges for a distance of nearly 8 miles. They consist of shales, sandstones, and conglomerates with calcareous bands in the Knoxville beds. There are no indications of the existence of heav,y deposits of eruptive material or other forms of rapid deposition, and the faulting and folding is slight and local. Doctor Stanton, who has made later and more special examiaations, thinks, however, that certain facts oliserved l)y him may idtimately somewhat modify these extreme results. There can l)e no doubt that those beds embrace practically the whole of the Lower Cretaceous, and Doctor Stanton would place the Chico beds in the Upper Cretaceous, witli their lower portion on about the horizon of the Cenomanian. These collections reached Washington in the autumn of 1895. I made a preliminaiy study of them during the winter, comparing them carefully with those previously received from substantially the same localities. In many cases they proved much fuller and greatly increased our knowledge of the flora of those beds, but in others, as already stated, Mr. Storrs and I were unable to find any additional material, and in still others what we obtained was inferior in cjuantity and quality to that previous!}' collected. Professor Fontaine was at that time engaged on other work and could not undertake the determination of these collections. I had planned the series of papers now in progress and, as stated in the first paper, deemed it important to begin witli the lowest Mesozoic floras and work up to those of the Lower Cretaceous. This plan has l^een adhered to, although it necessitated the postponement of the determina- tion of collections made earher than some that have l)een published. The history of this work has all been fully given in its proper place. "Am. Jour. Sci., 3d ser., \o]. XL, December, 1890, p. 476; 15ull. (Iim.I. .Sue. America, Vol. \", 1S'.H, pp. 439-4-10. '' Bull. Geol. -Soc. Americu, Vol. V, 1894, p. 438 (footnote). FLOUA OF TlIK SHASTA KOUMATION. 217 No iimr, however, was lost in liavini; tlu> Ci'iMaccovis ])lants determined, and in Jnly, 1897, all the colKH-tions thus far made from the Shasta beds, inchidino; those tliat had already l^een once studied and report(Hl u])on, wei'e s(>iit to Professoi' Fontaine and h(> made a Ihoi-ou^h examination of it all in the lijjht of the latest collect ions. His final report was com- pleted and ti-ansnutled on Fehruaiy 121, ISOS, and tlie collections wei-e i-elurn(>d to Washin<>;ton. Owing to the n(>cessity of having the .lurassic fioi-as illusti'ated first, I did not send the types of the Shasta fioi'a to the division of illustrations till near the end of May, 1S9!). Th(\v were promptly taken up and completed in July. Dui'in.ii; this time a few small collections of HorscMown j)lants had Ix'en made fi'om several localities in Oregon. A single specimen was obtained by Mr. Dillei' from Mr. Claude Rice, who gave the locality as 25 miles a little south of east of Buck Mountain. It consisted of a small dicotyledonous leaf on a rock (;ontaining shells of Horsetown age, and was fii'st sent to Doctor Stanton, who turned it over to me on April 25, 1898. Three other specimens came into n\y hands through Mr. Diller on May 31, 1898. One of them, in two complementary parts, was col- lected by Mr. Rice, but the precise locality is not stated. The other two, one of which was also in tw'o complementary parts, were collected by Mr. Brown on Cow Creek at Riddles. Another specimen, also collected by Mr. Brown, but from a different locality, viz, on Iron Mountain Creek, half a mile above its junction with Cow Creek, was sent me on February 9, 1899. This locality is in the Knoxville beds and is only about 3 miles below Nichols station, where the collection of Jurassic plants was made. I visited it in c o npany with Mr. Brown on September 18, 1899, l)ut we could no find more plant impressions. Aucella is very abundant in the same rocks that yielded the plant. On September 21-23, 1899, I visited several of the localities near Riddles where Mr. Brown had found fossil plants in the shell-bearing shales of the Horsetown beds, but I was mainly unsuccessful in finding plants, which are very rare. The interest aroused by bi'inging to light the specim'^n colbcted near Pettyjohn's ranch by I\Tr. Gilbert Thompson in 1882 led to a renewed effort to rediscover the locality. Mr. Thompson furnished Doctor Stanton with full details, including a sketch map of the region, and on September 15 and Ki, 1892, the latter visited the place and made a very careful examination of the beds. He collected specimens showing 218 MESOZOIC FLOKAS OF UNlTP:i) STATES. obsciire vegetable impressions at five points, designated on his labels as follows in their i-elation to Pettyjohn's ranch: 1, about 3 miles a little west of north: 2, about '.i\ miles a little west of north; 3, 2 miles below (southeast?); 4, li miles north; 5, l-i miles northeast. Nos. 3 and 4 are on the Cold Fork of Cottonwood Creek, No. 4 being in the canyon; No. 5 is on the trail. None of the specimens bear any close resemblance to that obtained by Mr. Thompson, and the presumption is that the exact locality was not found. These specimens were sent to Professor Fontaine on Feljruary 16, 1903, and his report upon them was received on March 10. It is as follows : I have carefully examined the fossil plants collected by Dr. Stanton from the vicinity of Pettyjohn's ranch, Tehama County, Cal. The collections unfortunately show nothing that can be positively determined. Most of the specimens can not be even generically determined. Many of them are vague imprints of stems or small scraps of leaves with none of the original form preserved. The best specimens do not show enough to give any idea of the true character of the plants. The following are the collections now in question : Collection No. 1, as designated above. — This was obtained from a locality about 3 miles a little west of north from Pettyjohn's rancli. It contains 6 specimens, none of which are even approximately determinable. The most that can be made out is that they are fragments of plants. Collection No. 2. — This comes from about .SJ miles northwest of Pettyjohn's ranch, several hundred feet above No. 1. It has only :> specimens. Onlj^ 1 of these can be even approximately determined. It is the basal portion of a dicotyledonous leaf that resembles Celastrophyllum hroolcense Font., from the Aquia Creek beds of the Lower Potomac of Virginia. It resembles also Ficus atavina Heer, from the Atane beds of Greenland. There is not enough of the fo.ssil to determine its true character. Collection No. 3. — This collection has 12 specimens. None of these seem to be dicotyledons. The fossils come from the Cold Fork of Cottonwood Creek, about 2 mUes below Pettyjohn's ranch. Seven of the specimens show only vague scraps of plants. Five of them have each a small portion of an ultimate pinna of some fern that shows onlj^ a few poorly preserved pinnules of the type of Thyrsopteris rariner- vis Font, or Asplenium Diclsonianum Heer. They all apparently belong to the same species. It may be either of the two forms mentioned above, for the mate- rial does not suffice to determine the ([uestion. The former of these occurs in the Lower Potomac and the latter ranges from the Kome strata of Greenland to the Atane of the same region. Collection No. 4. — This comes from the canyon of the Cold Fork of Cottonwood Creek, li miles north of Pettyjohn's ranch. It is the largest of the collections and FLORA OF TlIF SHASTA FORMATION. 219 font ains .'50 specimens. Nineteen of tiieni show not liiiii: lliatcan be even a|)])i()xi- uiately detennined. Most of these have vague imprints of stems. The remaining s|)ecimens contain fragments of i(>aves that give some hint of their character, but none of them can l)e [xisitively determined. 1. .\ fragment of a rather hirge leaf, sliowing no principal nerve' The iier\a- tion. whicii is poorly preserved, is all e(|iially strong and sliows an anastomosis that may be either tha