Bygenm S i a a * ge oe ee



Founded by private subscription, in 1861.


No. IGS4,



A Wa aaa I.—FOSSILS from the base of the Trenton Group. By J. W. Sanam, bk Palzontologist of the Geological Survey of the United Kingdom. if: Sk

= IL—GRAPTOLITIDE# from ihe Lower Silurian. By Professor Tasces ts Hatt, Paleontologist of the State of New York.

TIL—CYSTIDE and ASTERIDEA. By E. Bituines. Genus CYCLO- Pe CYSTOIDES. By H. Brtines and J. W. Sautmr. Paleozoic ae Bivalve ENTOMOSTRACA. By T. R. Jones, Esq.,F.G.8.

TV.—CRINOIDE from the Lower Silurian. By E. Biturnes.

Decades I. and III. are published. Decade IV. will be published on the Ist of J une, ; and Penne II. in about six months.


B. DAWSON & SON, 22 Great St. James Street, Montreal; A. H. ARMOUR & oe yt Toronto; JOHN WILEY, New York; HECTOR BOSSANGE & SON, Paris, Franters SAMPSON, LOW, SON & Co. Psuden, England.

Montreal, Ist May, 1859.

; : O = fC <a) RRS ys 8 Say pA = eye SSR 1 = 000 9 8 eae) SA Sy = N= KA =a 28) |B ee


A eee a <a : = Z = ae - 6 28 S i S 5 ss ps : a a a ee = = ao 5 | S = a= ic cg i a | nn he ef aan, < + ae pa | ee a ms = ; ae | Z a 65 << Hx Dn FES <a =, QA = Z, Re I = ea ~ & 5 SS S 2 ea : <a i we E Pe Zi = = 2 2 | : fo me) 4 ida’ —_— a iz ra = 3 mee = : | | el = <q ly i a) ree BA]

Pibrary of the Museum COMPARATIVE ZOOLOGY,


Founded by private subscription, in 1861.

No. J G38, UY












7 es ad | Wa tie

; Re

Reine FA OC i.

The Graptolites of the Quebec group, which are the subject of the present Decade, were first discovered at Point Lévis in 1854, and were then confided for investigation and description to Professor JAMES Haut of Albany, the distinguished Palzeontologist of the Geological Survey of the State of New York. This was prior tu the appointment of Mr. E. BILiines as Palzeontologist to the Canadian Survey, in 1856. After a preliminary notice communicated by Professor Haut in 1855, extensive additions were made to the collection of Canadian Graptolites, which were placed in his hands; and in 1858 descriptions by him of nearly all the species here figured were published in the Report of Progress of the Geological Survey of Canada for 1857, but without illustrations. Figures of one of the species were however published in the Canadian Naturalist for June 1858; but various accidental difficulties having occurred in the preparation and engraving of the plates, the publication of the Decade has been delayed until the present time. |

It is to be remarked that although the name of Decade, under which Parts I, I11,,and Iv were published, is still retained, this monograph is illustrated by not less than twenty-three plates. These are all from excellent drawings by Mr. R. P. Wurrrietp of Albany. ‘Twenty-one of them were engraved on steel by Mr. James Dutuim of New York, and the remaining two were lithographed by Mr. F. J. Swinton of Albany.

W. EH. LOGAN. GEOLOGICAL Survey OFFICE, Montreal, December 1864.

RT li ae ta

Ae et

\ ‘me? ‘J din}: 4, od A eh

if iamenint rn Bits Ng



§ I. Nature and form of Graptolites; formerly known only in fragments; regarded by early writers as of vegetable origin; their reference to Cephalopoda by Wahlen- berg, Schlotheim, Geinitz, and Quenstedt, page 5.—Their first reference to Poly- piaria; the supposed affinity of Graptolites to the larve of Echinoderms, 6.—The name Graptolithus first established, 7—What forms are properly referable to the genus Graptolithus, 7-8.—Dendrograptus, its mode of growth, and gradations through Callograptus to Dictyonema, 10-12.—Forms of some Graptolites similar to the recent genus Crisia, 13.—Forms with diverging stipes, 14.—Doubly serrate forms (Diplograptus), 15.—Phyllograptus, Retiograptus, Retiolites, Rastrites, 16. Thamnograptus ; Inocaulis, and its affinities to Graptolites, 18.

§ II. The central or basal portions of Graptolites, 19.—The radicle or initial point, 19. —The funicle, 19.—The central disc, 20.

§ III. Nature and parts of the stipe, 21.—The solid axis, 21—The common canal, 23. —The cells, cellules, or calycles ; their form and mode of development, 24.—Their similarity to the recent genera of Hydroide, 25.—Cell-walls, their double appear- ance and structure, 27.—EHxcavated cells, 27.—Their internal structure, 28.—Cells in Retiolites. Apertures of cells in some species of Graptolites; in Dendrograp- tus, 29.—The cells of Dictyonema, 29, 30.—Ornaments of the test in different genera and species, 30, 32.

§ IV. Mode of reproduction and development in Graptolites, when first made known, 32.—Different forms of germs, and their reference to species, 34.—Reasons for sup- posing that the diprionidian forms always existed as simple stipes, 36.—Develop- ment of cells in different genera, 37.

§ V. Mode of existence, 38.—Simple or compound forms; floating or fixed forms, 39.

§ VI. General characters of the Graptolitide, 40.—The condition of specimens heretofore described. Genera and sub-genera proposed : reasons for and against retaining these, 40 —43.—Diplograptus, to what forms the term is properly applied, 44.—Peculiar characters of D. putillus, 44.—Biserrate forms which are properly separable from Diplograptus, 45.—The genus Retiolites, 46.—Differences observed between Huro- pean and American specimens, 47.—Retiograptus and other genera considered, 47-50.—Synopsis of the genera of Graptolitide, 50.

§ VII. Geological and geographical distribution of the Graptolitide in the rocks of Canada and the United States. Equivalency of the American graptolitic beds with those of Europe, 51.—Mr. Barrande’s opinion of the age of the graptolitic schists in Northern Europe and in Bohemia, The graptolitic beds in New York. Discovery



of a Graptolite in the Potsdam sandstone. The graptolitic beds of Canada, 52

Graptolites of the Utica slate and the Hudson-River formation, 53—54.—Table showing the vertical distribution of the genera of the family of Graptolitide, 55.— Table showing the geological distribution of the species of Graptolitide in Canada

and the United States, 56-58.

§ VIII. Historical notice of the genus Graptolithus, 59 - 64.


§ I. Synopsis of the species of Graptolitide from the Quebec group, described in this

Memoir, 65 - 68.

§ Il. Descriptions of the species of Graptolitide illustrated in this Memoir, under the

following genera:

GRAPTOLMEHUS .«e\e!ajs\k.c/s cles sis/aie sieve ue here mis Geka aust toveiterenrie ale tevelsy«' ais (leis(e(e[staie aes TET OGRA DUS tors jciscoxey els 010) cjereiercle clove a ayeeroRete RAG Hes. Anwae sae » o.eieieere tee CLIMACOGRAPTUS ....... ehicreleteierodete ae Apusiere ale deeystee © aie ete aie Se Lit IEA NO TAT rn) as, 0) 016) o's, nile, s\e}/oia' ei eedetors Bisteteravarslekere ais 4 c aa acie sie 's omni ltcs EURO EURIAIP TTS cle fo\fello] oat ctellsiia ays) ets wb ohayayeter onsleneteltors Salad aes ei sitoyseatnnes Soke Srarerctage 115 PAVTHOGRAPTUS) «\0jcis «ie /s\e\e/s) « aVn vais ralohotainiolalelcfeieiavcisieiavets, cetstelone MeeesHaoas Ll DENDR.O GRABINUA ttc mysimicraieielerclesielcrielele Arete lens ole leneta aie tiaras catatralstorens SionekS - 126 CATTOGRAP TUS). cic.0 sciceicie ste wets e ateastehacels ois tofe siete tele vciavers ots SU Sisieiewiate eels DT OMMONE MWA <psie!s:c10oredeteseiei ina stess.sieie sieeve : my lisriva:tureteretarete e eieiate ec efalarotammlents DIG OGRAP TUBS. ie citisleeisafoieiere = o's sveresejasvaustos SAR Ashen stearelbusls ae cee St KIMPAMNOGRIAE TUSiertater-.clevele ere es ule teretarevetnie etarclalate ciate: ateraycharcvaloh otevereneroraiteitelaletenmel 0


Descriptions of Graptolites from the Utica slate, introduced for comparison and illus-

tration, 143-147.

RINTDTI Ne tisic is Osrows love iaceteiniovel ofatevaleleve/ave laleneliaga tenes rates ateletempetereieiuietale ala\tesajetufarmials (ere etalereteterets


APPIMAVINEISS “(ola cs a's roreieh casera vejulotalaye tare fevereltiloleveleilein iaieha‘oteve feratavelaie,slavarereieicis cteiterelere reve sranaratn ts tate mL Ene


Page 146, 6th line from bottom, and page 147, 4th line from bottom, for Rurxo- GRAPTUS read RETIOGRAPTUS.

Page 147, second line from bottom, read The frond four times enlarged.”




Until recently the graptolites were, with two or three exceptions, known only as simple, straight, or slightly curving linear stipes or stems, usually lying in the same plane upon the slaty lammee in which they were imbedded. Nearly all these were evidently fragmentary, and, though varying some- what in their proportions, rarely exhibited anything that could be regarded as the commencement or termination of their growth or development. These bodies, in their flattened condition, present a range of serratures either on one or on both sides of the stipe; and seldom preserve more of their substance than a carbonaceous or corneous film or test of extreme tenuity. Under more favorable circumstances, these serratures are dis- covered to indicate the apertures of cellules, symmetrically arranged in reference to each other, and to the axis of the linear stipe. Others show parallel entire margins, with transverse indentations across the central portion of the stipe. This appearance we now know to be due to the direction of the pressure upon the body exerted at right angles to the cellules, and which will be explained in the sequel.

The earliest opinion regarding these fossils was that they were of vege- table origin; and they have been thus considered by some authors even at a very late period. Subsequently, they were referred by Wahlenberg, and after him by Schlotheim, to the Cephalopoda, being regarded as extremely slender orthoceratites.. This opinion may have received sup- port from specimens in such condition as G. sealaris, where the indentations are limited on each side by a continuous margin; but in such as present a single or double series of marginal serratures, the analogy seems very remote. Professors Geinitz and Quenstedt advocated the same view at a much later date; though it has since been abandoned by these authors, from more extended investigations. |



Professor Nilsson first suggested that graptolites were Polypiaria, belong- ing to the family Ceratophyta. Dr. Beck of Copenhagen regarded them as belonging to the group Pennatulide, of which the Linnean Virgularia is the most nearly allied existing form. Sir Roderick Murchison has adopted this view of the relations of the graptolites, in his Silurian System.* General Portlock has fully recognized the graptolites as zoophytes, and has pointed out their analogy with Sertularia and Plumularia.

The relations of graptolites with the Cephalopoda had already been fully disproved by M. Barrande (in the first chapter of his Graptolites de Bohéme’’), before the abundant materials for the refutation were dis- covered in the graptolites of the Quebec group ; and most naturalists were already agreed in referring these bodies to the class of Polypi, to which they doubtless belong.

More recently, Mr. McCrady, of South Osea has published a paper on the Zoological Affinities of Graptolites,”’} in which he has endeavored to show the similarity of the graptolitic forms with the Echinoderm larva, as illustrated by Miiller. There is certainly much resemblance between the enlarged figures of that author, and some forms of graptolites in the shales of the Hudson River valley; while some of the figures with central discs have a more remote analogy with certain forms from the Quebec group. Some of the toothed rods of the Echinoderm larvee likewise bear a resemblance to the graptolites figured by Mr. Suess ;$ and there are still farther analogies pointed out by Mr. McCrady, which, however, may not be regarded as of equal value by the greater number of naturalists.

For my own part, although admitting the similarity of form and of some of the characteristics which were very kindly pointed out to me by Mr. McCrady, long before his publication, I cannot recognize the analogy sought to be demonstrated. The establishment of the fact that these toothlets or serratures are the extensions of true cellules, each one having an independent aperture, and communicating with a common canal, should offer convincing argument against these bodies being other than polyp- bearing skeletons. But, in following the extensive series of forms now presented to us, we have much evidence to show that some of these were attached to the bed of the ocean, or to other bodies; while the greater proportion of the species and genera appear to have never been attached to the sea-bottom.

It may not be easy to determine precisely the family to which these

* Silurian System, page 694; and letter of Dr. Beck, pp. 695-6.

+ “Remarks on the Zoological Affinities of the Graptolites, by John McCrady, made before the Eliot Society of Natural History of Charleston, S. C., at the meeting of July 15, 1857.” [Extract from the Proceedings, vol. i.]

t Naturwissenschaftliche Abhandlungen. Vierter Band. Tab. viii and ix.


graptolitic forms should be referred ; nor is it certain that the extensive series now presented can all properly be referred to a single family. General Portlock has suggested that these bodies may constitute ‘“ several genera belonging even to more than one order.”’* That they are true Polypi, I believe we shall be able to show, both from analogies already established by various authors, and also from their mode of development or reproduction as exhibited in some of the species.

The specimens which have usually been observed or represented are simple disconnected stipes, doubtless the dismembered or fragmentary portions of fronds, which, presenting in the different species great varieties of form and aspect when entire, are nevertheless composed of parts so similar that these fragments, though indicating specific differences, offer little clue to a knowledge of the entire form.

The name Graptolithus was established by Linnezeus in the first edition of his ** Systema Nature,” 1736, and applied by him to the straight or curved forms which are serrated (celluliferous) upon one side only, of which G. sagittarius has been regarded as the type.t The propriety of this term is more readily perceived in its application to the fragments of the stipes of monoprionidian forms than to the central portions of the body of the same. In the spirally-enrolled forms, or those with four or more stipes uniting in a central disc, as well as in the variously-branching forms, the analogy is not so perceptible.

Taking those species which, in the form of their cellules and in the separated fragments of the frond, would be referred to Graptolites proper, and tracing them, as we are now able to do in many species, to their perfect condition, we find a great variety of form and mode of growth. In the simplest of these we have two stipes diverging from a radicle, or initial point; and the parts remain so complete as to admit of no doubt

Fig. 1.



that this is the entire skeleton of the animal. ‘The cellules near the base of the stipe are not so fully developed ; while also those near the extremi- ties have not reached their full dimensions, and the last one is sometimes barely perceptible, or just assuming its form from the common body. These characteristics are perceptible in the figures upon plates i, 11, and i.

In the next stage we have four simple stipes diverging from an initial point, and all evidently entire, as shown in the development of the cellules.

* Geological Report on Londonderry, &c., p. 318. } I shall elsewhere endeavor to show that G. scalaris is a diprionidian form exhibit- ing only one margin.


Fig. 2.


In some species of this mode of growth, the bases of the stipes are united in a more or less expanded dise or cup of the same substance as the body of the graptolite. The form of this dise is shown on plates v, vi, Vii, vill, and ix, and also in the accompanying figure of Graptolithus Headi.

Fig, 4. Fig. 3.


In a farther development in the same direction, we have fronds with eight simple stipes, which may or may not be united in a central disc, as in the accompanying figure.


In Graptolithus Loganit we have numerous simple ‘stipes united in a central disc or cup; while in some specimens otherwise precisely similar, we have no remains of the disc. © In all these species the parts are disposed in a symmetrical and bilateral arrangement.

Figs. 5 and 6.

-GRraPtoLitaus Locant.

The stipes of this species do not bifurcate beyond the disc, and there


are no cellules below the last bifurcation. The number of stipes in different individuals varies from sixteen to twenty-five, so that this charactercannot be made of specific importance. In another similar species without a central disc, from the Hudson River formation, we have above forty stipes, which do not bifurcate, so far as known, beyond the commencement of the cellules.

Fig. 7.


The separated and broken stipes referred by me to Graptolithus sagit- tarius* are probably of this species, occurring as they do in great numbers in the same beds in which this was found.

In other species with a similar general arrangement of parts, the main stipes are frequently bifurcated ; the bifurcations beginning near the base, and continuing as far as the parts can be traced in the stone (fig. 8). In some of the species of this character the cellules begin near the base of the stipes, while in one species they are not known to exist except on the outer branchlets.

Thus far we trace these forms through what appear to be very natural stages in the progress of development of the parts, which are all constructed upon the same plan, presenting only natural, and we may almost say con- sequent modifications.

The character of stipes and cellules in all these is such that the sepa- rated fragments would afford no means of indicating whether the part belonged to a two, four, or eight-stiped species, or to those with numerous simple stipes, or with branching stipes, unless the fragment retained a bifurcation.

A variety of form is exhibited in the division termed Dendrograptus, in which we may conceive of the numerous stipes near the base becoming

* Paleontology of New York, vol. i, page 272, pl. 74, fig. 1.


Fig. 8.


with a portion enlarged.

conjoined into one strong stem, with the bifurcating branchlets spreading above, and this stem probably fixed in the soil. We then have a represen- tation of the typical forms of this genus, as in the accompanying figure, and as illustrated on plate xvii, figs. 8 and 9, of this memoir.

Fig. 9.


In this species, as shown in the enlargements of the branchlets, we have


a form of cellule similar to that in observed fragments of the ordinary species of graptolites. The cellules are very minute, and, from the frequent ramifications, this would probably always be recognized as a branching species.

Some of the forms of Dendrograptus have slender spreading branches, and less rigid stems than the typical species, but still retain the angular cellules, as in figs. 1 and 2 of plate xvii. From these we pass almost imperceptibly to the slender spreading forms which I have termed Callo- graptus, plate xix, in which there is, apparently, some slight modification in the form of the cellule ; and on the other hand, there is an almost insen- sible gradation to the Dictyonema, plate xx, in which the branches are connected by lateral bars, and the whole developed in a flabelliform or funnel-shaped. frond, with angular cellules on the inner margins of the branches. (Fig, 10, plate A.)


There are certain forms of graptolites, which, though possessing linear straight or slightly curving stipes and angular cellules, like the typical species, have yet a different aspect, and do not so naturally fall into the series. Among these we find Graptolithus divergens (fig. 11), where the bilateral relation of the parts is still shown, but the celluliferous stipes or branches are arranged on the two sides of a slender stipe or rachis, and diverge on each side from what appears to be the centre or initial point.

Different specimens show some slight variations of these characters, but not any essential differences.



Another form, which we know only in small individuals, is illustrated in the following figures, which remind one of some forms of the recent genus Crista.

Fig. 12. Fig. 13. Fig. 14. Fig. 15.

(These figures are enlarged to twice their natural size.)

Tn another form with similar angular cellules, we have the following illus- trations of the mode of occurrence of the species.

Fig. 16,



Fig. 17. Fig. 18.

$e we


We have still another group, presenting some differences, more particu- larly in the development and form of cellules, than in their general form ; and seeming in some species to unite the characters of those having a single range, with those having a double range of cellules. The simpler forms of this type are similar to the two-stiped forms of the first series ; but in all those, whether of two, four, or more stipes, the cellules are developed on the upper side, or that side opposite to the initial point. In species like G. sextans, G. divaricatus, and others of this type, the cellules are on the lower, or same side with the radicle.

Fig. 19.


and enlargement of cellules.

In the accompanying figure of G. divaricatus, the frond consists of two simple uniserrate stipes ; and the same is true of Gf. sextans, except that

* The fig. 18 was theoretically constructed, but has since been verified by the discovery of a specimen having the same form and arrangement of parts.


it is united at the base for the length of a single cellule. In G. furcatus the stipes are conjoined for a distance of two or three cellules above the base.

In G. ramosus, a8 shown in the following figure, the lower part of the stipe, for a considerable distance, has a range of cellules on each side, parallel with the axis ; and becoming bifurcate above, it presents two stipes or branches, each with a single range of cellules. All the species of this group have a peculiarity in the form of the cellules, which will be noticed hereafter.

Fig. 20.



These species, in their mode of growth alone, present forms which might be regarded as intermediate between the mowoprionidian and diprionidian groups; though the typical forms, G. pristis and allied species, never show any tendency to a division of the parts of the stipe ; and we shall observe, as we progress, that these forms are connected with other differences of structure.

Fig. 21.


The species of this type (Diprion or Diplograptus) are simple linear or



sub-linear stipes, bearmg a range of cellules on each side, often showing an initial pomt orradicle at the base, and an extension of the slender axis above and beyond the celluliferous portion.

From these forms witha double series of cellules, we pass to the broad foluform stipes, which are apparently composed of four semi-elliptical parts conjoimed along their straight sides, and thus present four ranges of cellules. These forms (Phyllograptus) are, in some species, broad and short, while others are elongate, with sub-parallel sides, as shown in the accompanying figures.

[See also plates xv and xvi. ]

Fig. 22.


(See illustrations under generic descrip- a group of separated stipes as they lie upon tion, and plates xv and xvi.) the surface of the shale.

From the occurrence of a large number of these leaf-like stipes, some- times crowded together in a small space, I have inferred that they may have grown as Retiograptus. (Plate xiv, fig. 9.)

Following the forms with a double series of cellules, are those of similar general form, the Retiolites, which are known only as simple stipes. (Plate xiv, figs. 1-5.)

In the Retiograptus we have simple elongate stipes; and in one form (plate xiv, fig. 9) we have the stipes united by slender basal extensions in a spreading frond, in a manner not unlike some of the graptolites proper, with the parts in bilateral arrangement.

In the genus Rastrites of Barrande we find a departure from all of the preceding forms in the slender stipes with delicate ‘slender tubular


cellules.* The species of this genus, so far as known, have the cellules developed on one side only of the:stipes or branches ; though there seems no reason why we should not have species with cellules upon the two sides of the axis.

The species for which I have proposed the name Zhamnograptus con- sist of slender cylindrical stipes and branches, some of them very similar in general aspect to Rastrites, but the alternating branches are long and slender, and we have found no appearance of cellules on any part of the specimens known.

Tn the accompanying illustration the branchlets are given off alternately on the opposite sides of a stipe or rachis, and the slender solid axis can be traced from the main stipe into and along the centre of the branchlets. The analogy between these forms and the celluliferous graptolites of the preceding illustrations does not appear to be very intimate, but they occur in the same beds, and the fossil has the same texture and substance.

Fig. 24.


In the genus Ptilograptus (plate xxi) we have a somewhat analogous mode of growth, but there is always an aspect or expression of the fossil which distinguishes it from Zhamnograptus. In one species of Pizlo- graptus moreover, cell-apertures have been detected on one side of the branches or pinnulee.

In the Buthograptus we have a form bearing some analogies with the

* See illustration of Rastrites Barrandi under section iii, p. 26, of this introduction,


preceding, but the rachis is flexible, and is not known to be branched, while the slender alternating pinnule are flat and simple, as they have been observed in numerous individuals. ‘The specimen represented is even less curved than the usual condition of this species in the slaty Trenton limestone of Wisconsin.


Fig. 25.



Still more obscure, and perhaps remote in its analogies with graptolites, is the genus Jnocaulis, consisting of flattened scabrous stems, associated with Dictyonema in the shale of the Niagara formation, which, from their carbonaceous substance and apparent graptolitic texture, I have referred

to the Graptolitide.

Fig. 26. °




1. The radicle, or initial point. 2. The funicle, or non-celluliferous connecting portions of the stipe. 3. The central disc.

~ 1. The Radicle, or Initial Point.—In the most simple forms, or those with two stipes, as shown in all the figures on plate 1, there is a slender initial process, which I have termed the radicle. This presents a greater or less development in the different species ; in some being reduced to a mere pustule, or scarcely perteptible point, while in others it attains a quarter of an inch or more in length. Although in none of the species with a single range of cellules does this part show absolute evidence of having been attached to any other substance at the maturity of the fronds, yet it is possible that in the earlier period of its growth, the body may have been temporarily attached at this point to the sea-bottom or to some object ; though all the evidence is opposed to this view.

In some of the bi-celluliferous forms, and probably in all of them, there is a somewhat similar extension below the base of the celluliferous portion of the stipe, though it is usually more slender; but whether this is always the true initial pomt of the whole body, or whether it is only the broken point of attachment to a frond, may sometimes admit of doubt. It is con- spicuous in Phyllograptus typus; and we observe this feature also in Retiograptus ; but in one species of this we learn that it is only a broken process of attachment of the individual stipe, which existed as one of the members of the entire frond, the true initial point of which would be in the centre of the whole. (Pl. xiv, fig. 9.)

In all the forms of Graptolitidee which appear to have been free, the initial pomt or radicle is in fact the commencement of the solid avis, which will be noticed farther on. In those graptolites with two simple stipes, the little radicle-like process enlarges above, and the stipes, diverging in opposite directions, are closely united at their bases, and the cellules begin almost in the axil between the two. (Plate i, figs. 1, 3, 7, 9, and 10.)

2. The Funicle—In the graptolites with four stipes, the condition appears like that of two individuals of the two-stiped forms, conjomed by a straight connecting process of greater or less extent, with the radicle point in the centre, though often obscurely marked. This connecting process is always destitute of cellules ; and this, with its divisions, I have termed the funicle. .

In those forms with eight stipes the funicle is twice divided at its two extremities; and where there is a greater number of simple stipes, it is


correspondingly subdivided. Neither the central portion, nor any of its subdivisions, becomes celluliferous ; and these parts are not termed stipes or branches, according to the views I have entertained. It is only beyond the last subdivisions of this part of the body, as in G. Logant, that the celluliferous parts, or the true stipes, commence.

In one of the proper branching forms, however, the cellules begin immediately beyond the first subdivisions of the funicle, as in the four- stiped species. (Fig. 27.)

GRapPtoLtitaus Minzsi.

These barren, or non-celluliferous portions of the graptolitie body, are not otherwise essentially different from other parts of the stipe. In the absence of cellules they are consequently more cylindrical, and apparently more solid, asif the test were thicker, and the common canal less developed than in the other parts of its extent.

3. The Central Dise.—In several of the species having four simple stipes, in one species with eight, and in another with a larger number of simple stipes proceeding from a common centre, we find their bases united by a thickened corneous expansion of the same substance as the body of the graptolite. This appears to be composed of two laminz, which, at least in the central portions, are not conjoined, and the space-is probably occupied by some softer portion of the animal body. (Plates v-ix.)

The substance of the disc sometimes extends along the margins of the stipes, producing an alation, as in G. alatus. (Pl. vi, fig. 9.)

This arrangement of the parts of the body seems obviously adapted to give strength and support to the bases of the stipes ; but: beyond this it probably serves other purposes of the animal economy. In several specimens of G. bicornis there is a disc or bulb at the base of the stipe, which, spreading between the two oblique curving processes, envelopes, in the compressed condition of the specimens, some of the celluliferous part


of the stipe at its base. In other specimens we have a crescent-shaped extension, as if the disc were in process of development, or perhaps of absorption. Much the larger proportion of the specimens of this species, however, are destitute of disc or bulb, and have every appearance of being complete without this appendage.

These aspects of the species are shown on plate A, figs. 18, 15, 16, and 17.

How far the bulb-like appearance at the base of some of the species of Dendrograptus may correspond to the disc of G. bicornis, I have not at this time the means of satisfactory determination.


1. The solid axis.

2. The common canal,

3. The calycles or cellules.

4, Nature and ornaments of the test.

1. The Solid Avis.—All the graptolites proper have been found to be provided with a slender solid axis,* while this feature has not been satis- factorily proved in regard to Dictyonema, and to some other forms.

In those species having a single series of cellules, this axis is upon the back of the stipe, or on the side opposite to the celluliferous margin; and in the branching forms it follows all the ramifications. In all the speci- mens where it has been observed, it is a slender cylindrical or flattened filiform solid body. In some extremely compressed specimens, this axis appears as a slender elevated ridge along the back of the stipe; and where the substance of the body has been removed, it leaves a narrow groove along the margin of the impression.

In the examination of large numbers of specimens of the monoprionidian species, we have never found the axis prolonged beyond, or denuded of, the cellules; as shown in G. colonus, by Barrande, in his Graptolites of Bohemia. (Plate ii, fig. 5, of that memoir.)

In all the specimens where the extremities of the stipes are entire, as represented in plates i, ii, and iii of this memoir, there is never any extension of the axis beyond the last partially developed cellule; and the number of specimens in this condition is considerable.

Tn the graptolites with two series of cellules, the solid axis is very

* In those species with a single series of cellules, M. Barrande has ascertained that this axis is solid and cylindrical, its diameter not exceeding 4 millimetre, and its struc- ture apparently fibrous. (Graptolites de Bohéme, page 4.)



frequently seen extending beyond the celluliferous portion of the stipe at _its outer extremity ; while the radicle appears like the continuation of the same below the base. ‘The axis thus appears to be the foundation on which the other parts are erected. In those specimens, however, which present so great an extension of the solid axis beyond the stipe, the cellules may have been removed by subsequent causes.

I am able to corroborate to some extent the observations of M. Barrande in regard to the apparent double character of this axis. In some ex- tremely compressed specimens it is marked by a longitudinal groove or line of division ;* while in others, a double impression has been left by the removal of the substance.

In some specimens, particularly the younger ones, the solid axis has been seen extending beyond the base of the stipe, as a duplicate process, exhibiting a character as of a double radicle. In some solid specimens of one species, where the tube had been filled with calcareous mud, I am able to detect only a single round point; and a longitudinal section of the same species presents a slender filiform axis. It may be, however, that the parts are so minute and so closely united, as to render them undis- tinguishable.

In another species, with two rows of cellules, and in which the latter are of very different form from the preceding, the solid axis is a thin flat apparently double plate, extending across the entire transverse diameter of the tube, which is more than two-thirds as great as its longer diameter. The place of the axis is marked by a longitudinal groove on each side, not in a direct line, but slightly undulating to correspond with the cellules. M. Barrande conceives that the joining of the two plates of this axis may leave a very flat intermediate tube ; and in our specimen, there is appa- rently an extremely narrow space between the two. He farther supposes that each of these plates, composing the double axis, is separable, by decomposition, into two lamine, as illustrated in plate ii, fig. 3, of the work already cited.

The entire appearance of the species (plate A, fig. 10) is that of two monoprionidian stipes joined together at the back, the line of junction being indicated by the groove.

In one species of fetiolites there is a strong excentric or sub-exterior axis, which is nearly direct ; and in the same individual there is another undulating axis, to which the cell-divisions of one side are attached. In the Retiolites of the Quebec group, one side of the stipe shows a very distinct axis, while upon the other side it is very obscure.

In Retiograptus we have a very distinct central axis projecting below

* The aspect presented by the axis, when marked by a longitudinal groove, is pre- cisely that which a hollow cylindrical body would have if extremely compressed.


the celluliferous portion of the stipe, and, in one species, uniting in a funicle, and forming part of a compound form. In another species, the simple stipes present similar features, showing at one extremity the duplicate character of the axis.

In Phyllograptus, the central axis is apparently composed of four slender flat lamin ; but we have had no means of examining this part of the body in a satisfactory manner.

2. The Common Canal.—lIn all graptolites with a single series of cellules, there is, between the bases of these cellules proper and the solid axis on the back of the stipe, a continuous sub-cylindrical space or canal, which has been occupied by the body of the polyp, from which the buds, with their calycles forming the cellules, take their origin, and are thrown off at regular intervals. :

All the specimens which I have examined confirm this view; and in some of the species where the extremities are apparently entire, we observe the incipient development of the young cell from the common body. In those specimens filled or partially filled with the substance of the surround- ing rock, this canal is easily distinguished ; while in compressed specimens there is always a flattened space between the bases of the cell-partitions and the solid axis.

Tn those graptolites with two ranges of cellules, we have apparently a duplication of those with the single series, the two solid axes being jomed together, leaving a common canal or body on each side at the base of each series of cellules. If however the common body were thus divided, it would be by the solid axis becoming a flattened plate. This appears to be true of some species (as for example, plate A, fig. 10), while in others there is only a simple filiform axis visible. In this case, of course, there is not an entire division in the common canal after the manner of the other species. ‘This will appear farther on, under the illustrations of the structure of these bodies.

In Retiolites, the common body occupies the central portion of the stipe, giving origin to a series of buds on each side, while it is not divided by a central axis.

In some species the common body seems likewise to have more extensive functions ; for in such forms as G. divergens and G. gracilis, there is a long slender rachis, or tubular body, destitute of cellules except at its two extremities, and apparently consisting of a solid axis and a common canal, from which originate, at regular intervals, simple small stipes with solid axis, common canal, and cellules.

This appears to be one step farther in our knowledge of the origin or mode of development ; but it shows that a stipe or main axis may produce in one part celluliferous stipes, and in its extremities develop only cellules, as we see in the continuation of the main axis of G. gracilis and


G. divergens, where the continuation of this common rachis is marked by marginal cellules of the same character as those of the lateral stipes.

The common body therefore appears to perform the double function of developing the buds which elongate into stipes with cellules, and also of simple cellules; or we may consider the celluliferous extremity of the rachis, or main axis, as the termination of the reproductive process, or as analogous to one of the lateral celluliferous stipes. In the Sertula- rians we have something analogous to this mode of development. Some of the species have cellules along the common or main trunk, and produce at intervals branches or branchlets in place of cellules; others have a common body, or main stem, entirely destitute of cellules, but producing branches on each side (opposite or alternate, as the case may be), which branches produce cellules only.

So long as this rachis gives off only celluliferous stipes, it is analogous to those parts of the simple graptolites which I have termed the funicle, having within itself the sources of this development of the several parts. The mode of development differs from that of the branching forms, inas- much as the branches proper arise from a division of a cell-bearing stipe or branch, and of course a division of the solid axis and common body.

In Retiograptus, some specimens show the cell-divisions reaching nearly to the axis, leaving room on each side for a narrow common body or canal ; while in a species from the Utica slate, which presents one side of an entire frond, the cell-divisions of the exterior side all reach to the axis, leaving the common body on the inner or upper side. In a species from Norman’s Kill, near Albany, there are three parallel ranges of reticu- lations, with apparently two filiform solid axes, forming the divisions between the three meshes. This structure probably occupies one side of the stipe, while the common body may occupy the other side.

In Phyllograptus, the cell-partitions reach very far towards the centre, and the space left for the common canal is very small. We infer from the better-preserved specimens that there