1 • t » . 1 • * » • . V 4 • ' » , ' » , ' 1 i 4 • s f t 1 *. * • *-1 1 4 :,•/ . ^ l * f T . ^ ' • .* = r í • t • í ^ • # ■ . l y. 3 1 >' • ' í 1 f f • V 1 * • • f. I ^ ' 4 ' . ^ ; . > 4 '.V . ' I • ■ « • • . •, * 4 ' ■ ' i • ' ' ' • = i .r i - i •' * • ' . • • • - ^ ^ X ^ t * \ \ - h' ' » • » i í ' ' • • I * • . '. • • • V'.' V'-ï ' , .• ,• i': • ' ■ • ^ . • - • • íM . t i ' * • • ' , •• •. 4 > i ' • í . . • •• : : ' • • . ■ • • • I . • i , ' • % f • . . i * . . • r . i *' f , » • - . . ■ • ' V , *1 • .. • ' - 1 ■ ■ • _ •:** t ' < ' ' . ^ à ^ - ' ; • ' ."í * • ' ' J . - . » . ' ¿ i 4 • i • • ' . , : 0 4 ♦ « I 4 . é - • • - vï V/.0•"-'•■■.V ^ ••• /.».>•« , I • > . . • I 'i • . 1- . * < . / o ^^jCrx_A>Ox_ xJLA \TV_jaJ^^^3^LjL^ fi- cv A^. 1). 0 rYX__ rviLA3--Ox->ívxXrx'X^ CX/VV\^A/X_Í^ \5XJL_- rVv^uC5"Wx^^-ÍLb _ EYOLUTIOîi of the FAUNAS OF THE LOWER LTAS By ALPHEUS HYATT [From lilie Frockediîîgs of the Bostojí Socíety of Natural History Vol xxiv 1888.] PRINTED AT THE SALEM PRESS SALEM MASS July 1888 EVOLUTION OF THE PAUNAS OF THE LOWEE LTAS BY ALPHEUS HYATT [E-Yom the Proceedings of the Boston Society of Natural History Vol xxiv 1888,] FEINTED AT THE SALEM PRESS SALEM MASS 1888 EVOLUTION OF THE FAUNAS OF THE LOWER LIAS. BY ALPHEUS HYATT. This paper is a short abstract of a chapter on the geological and faunal relations of the Arietidae prepared for a fortlicorning memoir upon that family. According to my classification this group com¬ prises the Angulatus group, which I formerly placed In a distinct family, the genus Pslloceras, the normal forms generally supposed to belong to the Arietidie by all authorities, and also the remark¬ able genus Oxynoticeras, which I formerly classified as a separate family, and which is very generally referred to the Amaltheus group by other authors. The inclusion of so many important forms makes the Arietidse, from our point of view, so large tliat it con¬ tains the vast majority of all the lower Lias Ammonitinie, and it becomes necessarily the source, or ancestral focus for all the later formed groups of the Ammonitinie^ of the Jurassic in Pmrope. There are many more or less complete lists and monographs of local fau¬ nas in the province of Central Europe, and extensive collections which afford a solid basis for comparison. The preliminary work of Prof. Jules Marcou^ in synchronizing the minuter subdivisions of the Jura in Central Europe was completed by the more extensive application of the same principles by OppeU, who visited, studied and synchronized the faunas of the different localities and identi¬ fied the same beds in a large part of this province. The illustrated publications of Hauer^, Neuraayer^, Wähner®, Geyer"' and Herbich® have also thrown a strong light upon the peculiarities of the faunas of the eastern part of Europe, particularly the basin of tiie North- *The Ammonitinse do not include Lytoceras or Phylloceras, etc. Such forms have been placed by the author in a distinct sub-order, the Ly toceratinaj. ® Roches des Jura, p. 23, 162, 173 etc, ^Die Jura-Format. Eng. Fraiir. u. d. siidwestl. Deutschl. Württ. Jahresh,, vol. 12-U, 1856. • Die Cephal. a. d. Lias d. norostl. Alpen, Denksch. K. Akad. d. Wissensch., vol. 11. ® See note 4, p. 587. • Mojsis. et Neum. Bcitr. vols. 2,3, 4, 5. ^Ceph. Heirl. Schieb, Abli. k. k. geol. Reichsan., vol. 12. • Das Széklerland, Mitt. Jahrb. d. k. ungar. Anst., vol. 5, pt. 2. (17) Hj'att.] 18 [May 16, eastern Alps. All of these researches, and many others not men¬ tioned, have made still farther advances in the classification of the minuter subdivisions or beds practicable. The principles of geographic distribution first announced by Jules Marcou^ have also been vindicated by Neumayer^ who has defined the homozoic bands of life in the faunas of what he has denomi¬ nated the Mediterranean, Central European and Russian provinces. The province of the Mediterranean according to this author in¬ cludes the Jura in southern Portugal, southern and southeastern Spain, a part of southeastern France, the Alps in Italy and Aus¬ tria, t'he Carpathians, and the Balkans farther to the east. The province of Central Europe includes all the remainder of France and Germany, England, the lands about the Baltic, the neighbor¬ hood of Brünn, and Krakau, and perhaps the neighborhood of the Dobrudscha. The Russian province includes central Russia, Pets- chora land, Spitzbergen and Greenland. According to this author these comparatively well-known areas are parts of homozoic bands, which encircled the globe during the Jurassic period, respectively representing the boreal, temperate and tropical relations of the faunas now existing on its surface. All the forms of the Arietidae except Oxynoticeras have been traced by the united work of Quenstedt, Neumayer and myself in the fauna of Central Europe to Psiloceras 2^l<^'i^orbe and this species by Neumayer and the author has been united with its still more primitive congener, Psil. caUpliyUum^ in the Northeastern Alps. Neumayer took the ground that all of the normal forms of the Arietidae were derived from the more ancient though apparently contemporaneous fauna of the Northeastern Alps. I take great satisfaction in being able to corroborate this opinion, so far as the more important or radical series of the Arietidae are concerned, but propose a still farther improvement in these views by excepting cer¬ tain series from this law which I think originated in the basins of Central Europe. Before, however, this question can be considerelastic stages. Each series, as we have just said, has similar stages or periods of evolution in time. aiKl the entire group is similarly related. Haeckel has named these three in a general way, the epacmatic period of progress, the ac- inatic or fully developed period in the history of a series, and the paracmatic, or period of the evolution of degenerate forms. The author pointed out these phenomena and their agreements among Cephalopoda, in the same year that Haeckel published Ijis views, but did not propose an appropriate nomenclature. The successive faunas may now be similarly classified and ar¬ ranged to accord with the ontology of individual and the evolution of the group, and st3ded the Epacmatic faunas, the Acmatic faunas, and the Paracmatic faunas. The P]pacmatic faunas occupied the Planorbis, Caloceras and Angulatus beds, the Acmatic were in the Bucklandi beds, and the Paracmatic characterized the Tubercula- tus, Obtusus, Oxynotus and Raricostatus beds. So far as the Arietidse are concerned, there is no justification for any chronolo¬ gical subdivision of the Lower Lias faunas unless it be into three parts as indicated by these relations. Hyatt.] 22 [May 16, Neumayer and Wähner have shown the ancient character of the fauna of the Planorbis and Angulatus beds of the Lower Lias in the basin of the Northeastern Alps. The appearance in the same region of the very abundant fauna of the Trias, so admirably described by Mojsisovics in his ''Mediterranean Trias Fauna," makes it still more likely that the geological succession was more direct in this region than in an}^ other.^ The richness of the faunas of the lowest beds of the Lower Lias in the basin of the Northeastern Alps so completely described and beautifully illustrated by Neumayer^ and Wähner,^ that they are models for imitation, has added a large number of radical species to those few formerly known in Central Europe but has not added correspondingly to the progressive series. Neumayer states,^ that the radical species of Psil, had varieties in Northeast Alps approximating to some in Central Europe, though probably derived directly from Psil. caliphyllum a smooth species of Psil- oceras not found in Central Europe. 1 have had a similar expe¬ rience in studying the Psiloceratites of Central Europe, and also Cal, Jolinstoni and other Caloceratites of the same fauna. In these species, varieties are found having more complicated sutures like those of the same species in the Lower Lias of the Northeast Alps. The shells of this fauna are very generally characterized by liaving more complicated sutures, the outlines of the lobes and saddles re¬ sembling the species common in the Trias, and quite distinct frcftn the outlines of the sutures in the normal forms of the Arietidœ of the Lias in Central Europe. The sutures of the normal forms of the Arietidse in Central Eu¬ rope, and of the species originating in that fauna, have, as a rule, simpler outlines and are evidently more or less degenerate. This fact accords with the theory that they were derived from Psil, plan- orhe. The sutures of the Arietidm are more like those of the young of Psiloceras than those of adults of the same species. The shells of this genus, and in fact all of the Ammonitinae, have more com¬ plicated sutures in the adult than in their own young, Neumayer has also stated that in old age of Psil, caliphyllum the sutures de- 1 This remark has been sustained by the discovery by Clarke of a Ti iassic form, an Arcestes in the ilhaetio bed of the vicinity of Aachen-See (see Geo]. Verhalt, d. Ge^j. nordw. v. Aachen-See, by \V. B. Clarke, now of Johns HotJkius tJniversity. ) Unterst. Lias, Abh. k. k. geol. Keiclisan., vol. 7. ® Mojsis., et Neum. Beitr. vols. 2, 6, 4.5. * See V2, p. 43. 1888.1 23 [Hyatt. generate and then resemble those of the adult of Psil. planorbe. The sutures of PsiL planorbe var. levis are obviously intermediate between those of PsiL Calipliyllum and Caloceras. The sutures of the latter are also intermediate between those of Psiloceras and Vermiceras, so that a most perfect series exists in Central Europe between the more complicated outlines common in Psiloceras and Caloceras and those of Vermiceras, which are simpler and purely Arietian in outline. The sutures have been, therefore, to a certain extent, arrested in their development and evolution in the normal forms of the Arietidse and this is a marked characteristic entirely compatible with progression in other characters. In studying the genetic history of each series by itself, species after species and variety after variety, it has been found practicable to attempt to follow out the migration of forms on the same horizon from one fauna to another, and, though no doubt very imperfectly, to make a beginning in the direction of tracing the evolution of faunas chorologically. The results of this work are very sug¬ gestive, and it may be said with approximate accurac}^ that some faunas are autochthonous or give rise to new forms and, that oth¬ ers, which I have called residual faunas, are not favorable for the evolution of new species or varieties. It seems very probable that the basin of the Northeastern Alps was the place of origin for Psiloceras, Caloceras, Schlotheimia and Wsehneroceras. The migrants that peopled other basins were not the degenerate or highly progressive forms of these series, but often, and perhaps generally, the radical forms. Thus out of the exten¬ sive series of the Radical Stock, Psiloceras, and of Wsehneroceras of the Northeastern Alps, but very few species, and these almost wholly the radicals migrated to central Europe. The history is somewhat different in Schlotheimia, though even in this genus, there are many more forms in the basin of the Northeastern Alps than in Central Europe. There are undoubtedly certain facts, like the earlier appearance of Schlot, striatissima in the ''gelbe Sandstein'* of the Trias in Wiirtemhurg which do not coincide with the assumption of ori¬ gin of the Schlotheimian series in the basin of the Northeast Alps. Jt ought to be mentioned here also that PsiL planorbe var. levis ap¬ pears in tlie same basin according to Quenstedt in the Bone-bed, which is either the uppermost bed of the Triassic or the very ear¬ liest bed of the Lias. The existence of a bed at the base of the 2 Hyatt.] 24 fMay 16, Planorbis zone containing Psiloceras alone has not yet been demon¬ strated in the Northeastern Alps, though such a fauna exists in South Germany, Nevertheless the evidences palieozoölogically are so complete, that I feel disposed to support the opinion that the four series of the Arietidín above mentioned were derived by choro- logical migration from the Northeastern Alps. This conclusion must be true, or else the separate and distinct origin of the faunas of South Germany and the Northeastern Alps and perhaps others like that of England will have to be admitted. It is very difficult, if this last theory be accepted, to account for the phenomena subsequently occurring in the residual faunas. The residual faunas show that the radical types of series are apt to occur later in them than in those that have been called autochtho¬ nous. The greater completeness of the earlier faunas of the North¬ east Alps in radical forms and series also coincides with the similar completeness observed subsequently in the basins of South Ger¬ many and Cote D'Or in the forms of the progressive series, which alone appear to have arisen there. Thus in basins where forms and series must have arisen according to the evidence afforded by the earlier appearance of radical species, there are, as a rule, larger numbers of species and varieties than in those which do not possess radical species, or in which the}^ appear later in time. The hy¬ pothesis of the independent origin of series in these two principal autochthonous basins, would also involve the corollary that pre¬ cisely parallel series and identical species connected by similar va¬ rieties could also be independently evolved in these different ba¬ sins, The English basin also possesses very complete series and like South Germany has a bed exclusively occupied by PsiL pZan- orbe but the absence of peculiar forms, and the later incoming of radical species of other series, show that the hypothesis of the in¬ dependent origin of similar species or polygenesis cannot be ad¬ mitted without more evidence than is now in our possession. The future completion of the geological evidences by the discovery of still more complete series in the Planorbis, and later Triassic of the Northeastern Alps seems, therefore, to be likely since, as we have said above, tlie unexampled [)rorusion of the Triassic fauna and the progress of recent discoveries point also to the same re¬ sult. It seems preferable to adopt this conclusion as presenting the least difBculties. The view, that South German}^ might have been the basin in which the radical stock and all the radical series 1888.] 25 [Hyatt. of the Arietklae (except Vermiceras) originated, is directly con¬ trary to all the testimony of paleozoology, and has in its favor only the earlier occurrence in that basin of Psil. planorbe and Sehl, striatissima and catenata. Vermiceras has an independent history. The radical species of this genus, Cal. laqueam^ appeared in South Germany in the Cal- oceras bed and probably migrated thence to the Northeastern Alps appearing in the modified form of Verm, prœspiratissimum {sp. WäJiner) in the Angulatus bed of that fauna. This series also met with the most favorable opportunities for the evolution of new modi¬ fications in the faunas of South Germany and the Cote D'Or, The members of the genus found in the Mediterranean province are sim¬ ilar to the more degenerate forms of this genus and do not indicate the existence of very favorable surroundings in that basin during the Lower Lias so far as this series was concerned. The species of Arnioceras are not abundant or large in the North¬ eastern Alps basin, whereas they are very abundant and represented by large numbers of radical and progressive forms in the basin of South German}'' and the Cote D*Or. The radical species also oc¬ cur in the Angulatus bed of these basins earlier than that of the Northeastern Alps. The radical species of Coroniceras occur on the Angulatus ho¬ rizon in all three of the faunas just mentioned, but the evolution of the new forms of the series and two sub-series, and the pres¬ ence of transitional forms, are marked characteristics of the faunas in South Germany and Cote D'Or and it is probable that it origi¬ nated in these basins in the province of Central Europe. Agassiceras lœvigatus and striaries appeared simultaneously and were about equally developed in the basins of the Cote D'Or and Rhone, but have not been found so early as the Angulatus bed in any other faunas. It seems more likely from this and the abun¬ dance of other forms of the series that it appeared first and was best represented in these two basins. The evidence with regard to Asteroceras is more doubtful. It may have first appeared in South Germany with Aster, ohtusum in the Upper Bucklanki bed ; or, it may have been an exception to the general rule and appeared first in the residual fauna of Luxemburg where this species occurs on the same horizon. The collateral ev¬ idences, however, are in favor of South Germany. It has perhaps more remarkable degenerate forms in England than anywhere else. Hyatt.] 26 [May 16, Oxynoticeras made its appearance on about the same level in all the faunas so far as can be judged by the evidences ; but the other species of this series and of the sub-series of Oxyn. Greenoughi show that it probably arose in the Cote D'Or and was distributed from that centre. The basins of the Northeastern Alps in the Mediterranean prov¬ ince and of South Germany and the Cote D'Or in Central Europe may tlierefore be considered as marking out a zone during tlie time of the Lower Lias, running east and west, which was especially fa¬ vorable for the evolution of the new forms and series which are in¬ cluded in the famil}'" of the Arietidiie. I have accordingly named these basins the Zone of Autochthones, or zone of autochthonous faunas in the Lower Lias. The fauna of the Lower Lias in the basin of the Northeastern Alps is, however, not in the zone of autochthones after the deposition of the Angulatus bed. This zone, before the deposition of the Lower Bucklandi bed, had become narrowed in its easterly extension and was confined to the faunas of South Germany and the Cote D'Or- This is shown not onl}' by the diminution in the number of forms of each series represented in the strata of the basin of the North¬ eastern Alps, but also by the concentration of the beds and the mixed character of the faunas above the Angulatus bed. The best general summary of the facts in regard to the faunas of the different beds of the Lower Lias has been given by Wähner^ in his He- teropischen Diiferenzirung des Alpinen Lias." This author, whose keen discrimination of species I have had constant occasion to ad¬ mire, has been able to distinguish all the Oppelian beds from the Planorbis to the Angulatus bed in the Northeastern Alps, but he has entirely failed, as have others, in defining the beds above this, so as to make a parallel series with those of Central Europe. Stur,^ Giimbel,^ Geyer,^ Herbich,^ and others may be quoted in favor of the opinion that the faunas of the strata above the Angulatus bed are mixed, and that the beds cannot be separated as in the prov¬ ince of Central Europe. The}^ have failed in obtaining an}" direct evidence of subdivisions after careful investigations, and their ci¬ tations of the minute work of others show the same result. Some of these authors go so far as to claim that the faunas above the Angulatus beds in the Northeastern Alps not only contain species * Verb. d. k. k. geol. Reichsan., 1S88, p. 108. ^ Geol. d. Steinnark» p. 433. * Geogn. Beschrieb, d. bayeris. Alpen, p. 428. * See op. cit. ^ See op. cit. 1888.] 27 [Hyatt common to all of the higher beds of the Lower Lias, but also some species which occur only in the Middle and Upper Lias of the ba¬ sins farther to the west. So far as we have seen the forms, or noted the observations of others on the species of the Adnether and Hierlatz basins in the Northeastern Alps, this last extreme opinion does not seern to be sustained. That the faunas are mixed, as noted above, seems to be established, however, by the direct observations of several authors, notably Geyer and Herbich. These facts appear also to accord perfectly with the conclusions given above. If the Northeastern Alps were the seat of origin for the radical series during the time of deposition of the earlier beds, and the more western basins became the originating centres of the O o more progressive series, and, finally of the degradational series dur¬ ing Bucldandian and later times, then we can account for this con¬ fusion as due to return migrations from the western faunas towards the east. The radicals of the ditterent progressive series could then have arrived in the Northeastern Alps with sufficient rapidity of succession to have become mixed in the same rocky stratum ; and the later forms ma}*' even have overtaken the radicals while on their journey to the east and have arrived together in this basin, which would thus have become changed from an autochthonous to a re¬ sidual fauna. Another fact in favor of this view is the curious confirmation af¬ forded by Neumayer's remarkable conclusion that the faunas of the Mediterranean province are distinguished by the earlier ap¬ pearance and origin in them of the Lytoceratidie, a family which is characteristic of the Middle and Ui)per Lias in Central Europe. This opinion is well sustained by the later observations of Geyer at Ilierlatz, and especially by Herbich's discovery of an extensive series of these forms in the Lower Lias of Siebenbürgen. The rise of the ArietidiB is, in large part, more completely re¬ corded in the Northeastern Alps, than in any other basin, but the acme of its history was reached in the faunas of South Germany, and the Cote D*Or. The records of its declining series are about equally well given in the rocks of the Lower Lias in South Germany and Cote D'Or ; and in England there is also a very complete series, especially of what may be called the more degraded forms of As- teroceras, and of the first or oxynoturn sub-series of Oxynoticeras. The faunas of the Italian peninsula have not been completely ex- Hyatt.] 28 [May 16, plored, but so far as known, they belong to the same category as those of the Northeastern Alps, and are included in the Mediterra¬ nean province, as has been stated previously by Nenmayer. There is no evidence, so far as I know, which sustains the conclusion, that there was a migration of forms of the radical stock or of the radical species of the different series of the Arietidae from the south through the basin of the Rhone into South Germany. On the contrary, the basin of the Rhone contains a fauna, which is almost entirely residual and might have been derived by migration from South Germany and the Cote D'Or. The faunas to the north of the au¬ tochthonous zone of the Arietidae bear also very strong marks of liaving been derived from those of South Germany or Cote D*Or. Tims in Luxemburg, Hanover, and North Germany in general, there is a prevalence of the same succession in the beds and simi¬ lar faunas, which indicate a more exact parallelism with those of South German}' than has been found in the Northeastern Alps. The faunas are, however, less complete, both geologically and palaeo- zologically in the lowest beds of the Lower Lias, especially in the smoother varieties of PsiL planorhe. In other higher beds there is a paucity in the representation of the forms of the different series a prevalence of smaller-sized specimens, and a later appearance of the radical forms of the different series, which show the basins to have been peopled by migrants from the south, England also apparently derived its fauna from the south. The beds are presented in detail as in South Germany ; but, in spite of tiiis fact, there is in that fauna a generally later appearance of rad¬ ical species, and, in fact, with the exception of the more degraded species of the Asteroceraii series, the fauna does not present any forms, which can be said to have originated there. The shells are of large size in many species and the surroundings were evidently very favorable for the growth of the individual, though not for the evolution of new forms. Information with regard to the faunas of localities farther north, like those farther south than those mentioned above, is not exten¬ sive. So far as they go, the records exhibit more or less deficiency in the deposits of the Lower Lias, and the rocks do not contain rich faunas or earlier appearing radical species. The theory of migra¬ tion from west to east along a zone which passed to the north of the Alpsdiifers from the suggestion of Mojsisovics in his work ''Dol¬ omit-Riffe Süd-Tirols, etc.," that the fauna of Central Europe may have come from Italy by the way of the valley of the Rhone. The 1888.] 29 [Hyatt. faunas of the Italian rocks have been referred to the Mediterranean province by Neumayer and several other authors, and the mixed character of the faunas above the Angulatus beds has been noticed, especially by Geyer. There seems to be every reason for closely associating them throughout the Lower Lias with those of the O o Northeastern Alps, and all the specimens we have seen are either peculiar forms or have a Mediterranean facies. It has seemed, therefore, unlikely, that any connection existed during the time of the Lower Lias between the Central European and the Mediterranean province to the west of the Northeastern Alps basin. This opinion, established on purely palaiozoological grounds, receives confirmation from two different geological con¬ siderations. The Planorbis bed is either absent or very slightly and insufficiently presented in Italy according to all the works we have seen, and this is also the case along the shores of Provence. It is present in deposits reaching from the neighborhood of Toulon to Nice, according to M. Dieulefait, but does not contain any Am- monitinse. This author^ draws a boundary line between the rocks of the Lowest Lias in southern Provence or department of the Var and those of the valley of the Durance and department of the Basses Alpes. His researches show very clealy by means of the geology and palaeontology, that no connections probably existed between these closely approximated areas during the deposition of the Pia- norbis and Angulatus beds, and farther, that the upper beds of the Lower Lias are entirely wanting in the Mediterranean part of Provence, while they are well developed in the valley of the Dur¬ ance. This sustains the views here advanced that the migrations began during the time of deposition of the Planorbis or possibly some¬ what earlier beds and flowed out from the Northeastern Alps in two directions. One took the route to the south and west into the Ital¬ ian region, and, perhaps, farther westward into Spain, though of this last we have no proofs to offer. The other stream of migrants spread more directly westward into Bohemia and South Germany first, then into the Cote d'Or, but reached the Rhone and all faunas south of the autochthonous zone later. In the same way as re¬ gards England and the faunas north of this zone, they all, as we have said above, contain forms which show that they were estab¬ lished later in time than those of the basins of South Germany, or Cote d'Or. * Ann. (lea Sci. Geol., vol. I, pi. v. Hyatt.] 30 [May 16,1888. M. Dieulefait's observations also accord with my opinion, that the faunas of the south German basin during the time of the depo¬ sition of the Bucklandi bed made a return migration to the east, peo- pling the Northeastern Alps basin with progressive forms first, and then spread to the south and west into Italy. It would be difficult to account for the absence of the upper beds of the Lower Lias in the Mediterranean part of Provence or for the mixed character of the faunas of these same beds in Italy upon any other theory than direct connection with the similar faunas of the Northeastern Alps in the east and a corresponding isolation from the faunas of Central Europe to the west. I am well aware of the great difficulties under which one labors in writing an article of this kind so far awa}' from the localities in which the phenomena have been and are still continually being ob¬ served. The large amount of negative evidence which has to be O O taken into consideration in all the countries to the south and north of the autochthonous zone of the Arietidse will also rightfully create doubt in the minds of specialists. Nevertheless, the positive facts are closely and ver}-" remarkably in accord with the genetic clas¬ sification of the species as traced out graphically in lines spreading from a centre in PsiL iñcmorhe or caUphylluin^ and with the subse¬ quent cycle in the life-history of the individual and of the series to which the individual belongs. "Whether it will be ultimately demonstrated that all the faunas of the Lower Lias were derived by chorological migrations, and corre¬ sponding modifications, from radical forms which arose in the three autochthonous basins mentioned above cannot now be discussed. So far, however, as American forms are known to us, they have a distinctly European faciès.^ The geology of North America also supports the conclusion that the Jura was neither very completely represented by deposits nor rich in the number of forms evolved, two deficiencies that in Europe mark what have been called resid¬ ual faunas. ^ Caloceras Orioni from Chacapoyas in nortli Peru can only with difficulty be fepa- rated IVom the stouter varieties of Calocerai salinaHum described and figured by Waimer from theNortheasteru Alps, and a form of Cul. nodotianum lately received from Peru cannot be beparated from some varieties of that species in Central Europe. The South American province includes tlife Jura of tlie Argentine Kepublic, Chili, and Peru. This fauna besides a number of peculiar species is composed of a mixture of Central European and Mediterranean bpecies. The North American px'ovince includes California and Vancouver's l&land. Tliesé faunas over and above their own peculiar species are mixtures of European forms. [From the Proceedings of the Boston Society of Natural History, Vol. xv January ly, 1870.] REMARKS ON THE PORPHYRIES OE MARBLEIIEAI). $ BY A, HYATT. ^ ^ ^ l-N, 1L.X-. Prof. A. Hyatt made some remarks^n support of the theory advanced by Mr. Bouvé,'and exhibited a map of Marblehead Neck, made some years back by the aid of the Plane Table Map of the United States Coast Survey, and also largely from observations made by the class of 1871, of the Mass. Institute of Technology. The outlines of the porphyritic, granitoid, and micaceous rocks were pointed out, and the first named rocks more particularly de¬ scribed, The porphyries are the underlying rocks and occupy the 1876.] 221 [Hyatt. greater part of the Neck, the southern shore only and an area to the north, of a few acres, being occupied by the micaceous rocks referred to above. Both these and the porphyritlc rocks are overlaid by patches of coarse granite containing flesh-colored feldspar. The precise deri¬ vation of the granites could not be determined. When the map was made I supposed them to be volcanic products, and thought they had been derived from the same source as the vein rocks penetrating the Salem syenites. This conclusion, however, is untenable. The Salem syenites, which are so well known from their peculiar lithological characteristics, occupy a space of about fifteen square miles in the townships of Marblehead, Salem, and Swampscot. The whole of this series of rocks has been completely shattered by extensive eruptions from below. This is not only the most remarkable characteristic of the surface, as long since noticed by Professor Hitchcock, but is par¬ ticularly observable along the cliff exposures of the shore lines ; some of these, where the walls are perpendicular, show the original rocky mass split up into angular fragments from the size of a man's hat to those which are many yards in diameter. The fragments have not been injured by their violent separation, and if the vein¬ stone could be withdrawn they would fit together witli the most per¬ fect accuracy. The veins are filled with rock, which in some places, is a compact red feldspar, and in others of a syenitic or granitic char¬ acter, varying greatly in color and aspect. The Salem syenites are crystalline throughout. There, are, how¬ ever, indications that they may have been originally stratified de¬ posits, though this conclusion must at present be considered very doubtful, and is merely mentioned in order to attract attention to this point. It has become evident to me th&t these Salem syenites are older than the adjacent porphyries and mica slates, and therefore that their veinstones have no relation in point of age and cannot have been the source from wl^h the somewhat similar overlying gran¬ ites of Marblehead and Beverly were derived. In fact, so far as my observations go, the conclusion appears to be unavoidable that the Salem syenites are remnants of a much older series of rocks than those to which the porphyries belong. Their characteristics are in every way distinct from the adjoining granites of Beverly, Gloucester, and Peabody, and the veinstones by which they are literally reticu¬ lated, do not extend upward through any of these or of the interme¬ diate rocks, the porphyries and the mica schists of Marblehead. These last have been described as Huronian by Dr. Hunt, and so Hyatt.] 222 [January 19, mapped by Prof. C. H. Hitchcock, but whether the Salem syenites belong properly to the next oldest system of Hunt's series or to the Laurentian, seems at present doubtful. This matter, however, as well as the subject of the chemical changes of the porphyries, will, I hope, be fully investigated by one of the Assistants in the Society's Museum, Mr. W. O. Crosby, and fully reported upon at some future time. The porphyries appear to overlie the Salem syenites iinconform- ably, and together with them are cut by at least two series of dioritic dykes, one running nearly north and south, and the other in a north¬ westerly and south-easterly direction, if indeed any system can be eliminated from the confused lines, which intersect each other in every direction on the surface. The porphyries, though varying greatly in aspect and in composition, are nevertheless but one forma¬ tion, and derived from a vast conglomerate which appears in Lynn, Saugus, and Marblehead, and is reported to occur under the granites on the Beverly shore. The originally conglomerate nature of the entire deposit is inferred by extensive observations made by myself at Marblehead Neck, ami by my assistant, Mr. W. O. Crosby, in Saugus, and the general identity of the purely crystalline porphyries of Lynn with those of Marblehead Neck, which are undoubtedly merely altered conglomerates. In some localities it is possible to study the various phases of the changes which may take place in the original conglomerate within the circuit of a few yards. Thus at one point on the ocean side of Marblehead Neck, the variegated conglom¬ erate is altered to compact light colored fcbite in one direction, in another becomes a dark colored porphyry with crystals of feldspar. The change into the felsite is the most instructive, since here it is possible to trace the included pebble of dark colored, banded por¬ phyry through all stages until it becomes a mere spot in the light colored matrix. During this change the pebble disappears by some process by which the structure is altered from without, the centre being the last point to lose its distinctive coloring or structure. This, and the unaltered form of the pebbles or masses, would appear to militate against the supposition that such a series of changes could only take place in a plastic or semi-fluid mass. But whether this was the case or not, and whatever the condition may have been, the flict seems to me uîuiuestionable, after a review of this locality, that both a fehlte and a true porphyry were formed out of a conglomerate, without any perceptible change having been made in the form of the 1876.] 223 (Hyatt. contained pebbles. This is shown by some of the masses of the truly crystalline porphyry in which the pebbles have entirely disappeared in fractured surfaces, but show the outlines of their uncompressed forms upon the external weathered faces. That the conglomerate porphyries cannot have been derived from the adjacent masses of banded and crystalline porphyries is inferred from the fact that the materials of the coiiirlomerate are not identi- cal. The pebbles contained in them are evidently derived from some older porphyries, and are quite distinct. Besides this, the traces of pebbles may be seen upon the weathered surfaces of the crystalline porphyries and felsites, and their transformations traced back to their original condition in the conglomerate in many localities. The change of the pebbles into more or less lenticular masses, streaks, or bands, in the formation of the banded porphyries, is also very interesting. In this case the re-arrangement of the conglomerate, itself recomposed from older banded porphyries, takes place in a similar manner, but with certain distinctive character¬ istics. The material of the pebble is seen to be re-arranged, as it were, by the action of the matrix, into alternate bands of dark colored porphyry and white feldspar, marked here and there with imperfect crystals, the remnants of the centres of pebbles which have otherwise entirely disappeared. This re-arrangement pro¬ ceeds from without, so that the pebble eventually becomes a lentic¬ ular mass arranged in alternate laminœ. This would seem to be the direct product of pressure upon the mass, which would naturally produce the lenticular form, and lead, especially if a moderate amount of heat were applied, to the production of bands of feldspar. But if we examine the form which the laminae of the coarse, peb- ply matrix assume during deposition, we find that this lenticular form can be explained without bringing in the aid of pressure. These layers can be traced in many specimens. They are concave around the bases of the larger pebbles, straight or horizontal only at the middle part or zone around the centre, and become decidedly convex as they are heaped up on the upper half of the inclosed mass. The changes which take place first affect the lowermost and uppermost layers of the matrix, converting them into bands of feld¬ spar and dark amorphous porphyry. These form a lenticular figure surrounding the pebble and the zone of intermediate horizontal lay¬ ers, exactly as the lines of the eyelids surround the ball of the eye, supposing the corners of that organ to be filled with solid matter Hyatt.] 224 í January 19, representing the horizontal layers. The changes in the majority of cases follow this pattern, so that the included pebble becomes re¬ duced much faster in its vertical than its horizontal diameter, thus assuming a more ellipticaL and" flatter 5brm. The whole series of bands, which are thus seen.^o arise from above and below simulta¬ neously, approximate more and more to a horizontal line in ap¬ proaching the centre of each pebble until they actually do meet on one common level. ' • An infinite number of pebbles arranged with the longer axes in the planes of stratification, and.-undergoing such changes as these just described, would, by the intei'settion of their laminae, form the more or less concentric or continuous and irregular bands which are to be O found in what are called banded porphyries. Another form of porphyry is also found on the Neck in which the pebbles seem to be absolutely flattened out, and then to fuse or run together at their extremities, formins dark continuous streaks or O 7 Ö • bands. The precise mode of the formation of this kind I did not succeed in following out, and in fact attempted, with regard to the others, nothing more than what could be accomplished by the most direct visual observations unassisted by chemical analyses. Never¬ theless some curious f.icts can be observed in the merely mechanical phenomena attending these changes. It is exceedingly interesting to note that so great changes, as those described, could take place, and in a mass which must have been sufficiently plastic to permit of a continuous chemical reaction between the elements of the pebbles and those of the surrounding matrix, and yet not so plastic as to alter the contour of the pebbles. Also, that diflereiit kinds of rock, felsites, crystalline ami banded por[)hyries, were produced essentially from the same conglomerate, but that in all of these, while the chemi¬ cal and physical changes in the pebbles differed, the general facts remained, that in all eases the loose materials of the matrix exhibi¬ ted the metamorphosis first, and the pebbles more slowly, the changes in the latter proceeding concentrically always from without inward. This would seem to indicate that the plasticity of the ma¬ trix, if it was plastic, communicated itself very slowly, if at all, to the contained pebbles. mi 'NOXSMVAS H vrí Ny3i '■ Vy ^V\J 0 4 On" Reversions among the Ammonites. By Prof. A. IIyatt. In some remarks relatinir the origin of characteristics among o o o animals, Mr. Hyatt stated that he had recently discovered a series of reversionary characteristics among the Ammonites which might be considered worthy of exceptional consideration. These are the peculiar extensions of the pilaa (ribs so called) and of the intervening sulci, or lateral depressions, across the abdomen or external periphery of the shell, characteristics found especially in Mlcroceras planicof^ia and lailcosta. The genus which was founded upon this peculiarity and the (" Dlscoceratidae)" Arietidœ, having been recently subjected to a revision, certain similarities of a very remarkable kind were observed. It was found, that among the Arietidae, Coroniceras roùforme^ occurring in the Bucklandibett " of Oppel, Cor. nodoaum/xn the upper part of the same bed, Amm. Blrch 'd^ ]ust above this in the Tuberculatusbett" and Asteroceras oblusum^ still later in the *'Obtu- susbett," all exhibited toa greater or less degree the planicostan pilae on the abdomen during some stage of growth in certain individuals. In Coroniceras nodosum this is especially remarkable, and the con¬ trast between the young in those individuals which show this stage, and the adults, with keel channels and septa all so typically arie- tian in character, is very great. In all these species the planicos¬ tan stage appears only in a limited number of individuals in each sj)ecies, and is always succeeded in course of growth by the features ju>t described of keel, channels and septa, peculiar to the family of the Arietidm. In Ophloceras rarlcostatum^ however, the latest occurring species of the lower Lias which has the typical septa of this family, the planicostan stage is superseded in course of growth only by a keel, this species having no channels. Of course, in trying to account for the presence of this transient characteristic, one follows the family back to its lowest representa¬ tives. These may be said to be two species, Cnloceras torus and Arnioceras cuneiforme^ the former closely allied to Psiloceras psilo^ 1870.] 23 ^ [Hyatt. J- notum in its septa, and the latter alsolnltí external characteristics. None of the lower forms, however, display, so far as observed, the planicostan stage, though they occur earlier than the species which do exhibit this peculiarity. The planicostan abdomen, therefore, must either be a new characteristic suddenly interpolated in the growth of some inclividjï&,ls, or aVeversion to certain ancestral char¬ acteristics which have been discontinued for a time in the lower members of the family. The lower forms of the Arietidm, Caloceras torus and such species as Amm. nodotianum^ with which this species is closely allied, have septa that are similar to those of certain Triassic species, such as Amm. Brunneri'^ Amm. Battcni Strachey, which also resemble Psiloceras psUonotum in their septa and forms. The affinity, therefore, is doubly proved through the latter species, which is a contemporaneous form, and by direct comparison. Besides these there are other species, such as Amm. lœoidorsatus Hauer, and Clydonites quadrangulus Ilauer, which show us that the planicostan abdomen is by no means a new feature. Thus, though we cannot assert that the Arietidaî are directly traceable to species in the Trias having the planicostan abdomen, we can say that the family on its lower borders have affinities with Triassic species, and that the plani¬ costan abdomen is found in the Trias. It is probable, therefore, that the same modification, when it occurs in the higher Arietidae, after a certain interval of time is a reversionary feature. The young of Coroniceras nodomm^ Amm. Sauzeanus of D'Orbigny, is succeeded in the next bed, the " Tuberculatusbett," by a new form, Mlcroderoceras Birchii^ whose young arc entirely distinct in their mode of development from any of the Arietidag. They are at first very cylindrical and smooth, then two rows of tubercles are intro¬ duced; and sometimes, though rarely, a specimen occurs in which the planicostan abdomen is presented. The septa develop to a more complicated outline in a shorter time than any of the species which follow in the same series or any species among the Arietida\ Very similar to this in its adult ornaments and septa is Mlcroccras biferum; in fact, I was disposed to think them members of the same genus, until I became aware that a representative species, " Mirrod. Hehertii" existed in the middle Lias. This establishes a distinct series for Btrchii, and makes it necessary to employ a different name. 1 Haidinger's Abhand., Bb. 3. p. 23, pi. v, figs. 7-9. Mem. Geol. Survey of ludia, ^Stol., vol. V, pi. i, p. 69; pi. v, figs. 2, 3. Hyatt.] 24 [October 6, The series wHch we are now considering, has three other closelj allied species in the middle Lias. The first is Microceras laiicosta, ^ whose young are precisely similar in all respects to MicrocerasHferum^ though the adults differ consider¬ ably, the planicostan abdomen being brought out more distinctly in the adult stage than in hiferum. Associated with this species is Microce¬ ras crescensy whose septa in the young have the same characteristic outline and proportions as in the adult of Opliioceras raricostatum, though the whole form and external features identify it with Micro- ceras laticosta» Then there is Microceras arcigerens, whose septa in the young are like those of the compressed form of raricostatum at an earlier age, just before the minor lobes and cells attain a decided prominence.^ The whorls in this species are flattened dorso-abdom- inally. The dorsum is broader than the abdomen, and this, together with the flattened aspect of the whorls and the early development of the closely set pilas, gives an umbilicus closely simulating that of raricostatum. The genus Androgynoceras returns to the peculiar pilae and tuber¬ cles of Microderoceras Birchii in the adult, though retaining the adult characteristic of Microceras until a late stage of its growth. This is especially remarkable in Androgynoceras hyhidum (D*Orb.), but becomes confined to an earlier stage in Androgynoceras appressum. The next genus of this same genetic series exhibits in Liparoceras indecisum the planicostan abdomen not later than the fourth whorl. In Liparoceras Henleyi this is apparent at an earlier stage only, and in Liparoceras Bechei it is absent altogether. The same mode of growth is here returned to, which was first ob¬ served in Microderoceras Birchii'y namely, a smooth, round whorl, suc¬ ceeded immediately by two lines of tubercles or spines, erected upon piliE which do not cross the abdomen, except as fine, distinct linear ridges. The difference between the two species, in other respects, is very great, sufficient, in fact, to constitute very distinct genera. It will be observed that we have here a closed series, one in which the 1 In the Bulletin of the Museum of Comparative Zoology, Ko, 5, this species ap¬ pears under the names of Microceras sinuosum and Microceras maculatum, two species which 1 now regard as the compressed and gibbous forms of Mic, laticosta, ® This compressed variety is the one figured by Sowerby, and can only be doubt-. iuUy referred to the same species as 0. raricostatuSf which is much flatter on the Abdomen, and altogether difierent in form as well as smaller, though precisely sim¬ ilar in the septa. 1870.1 25 [Hyatt. genetic connection is traceable from species to species, and these species agreeing quite closely, even as regards the two most widely separable forms in the proportions and outline of their septa. The planicostan abdomen is a reversionary feature, occurring transiently and rarely in Microderoceras Birchii^ but becoming characteristic of the adult in Microceras Viferum^ and the prominent peculiarity of the remaining forms of this genus. That this is not an artificial arrangement may be seen by consult¬ ing the geological succession of the groups. Microderoceras Birchii is found in the " Tuberculatusbett " of Oppel. Microceras hiferum and Microceras laticosta'^ in the " Oxynotusbett,'* the latter, how¬ ever, lasting into the middle Lias. In this formation it overlaps Androgynoceras^ which appears in the " Jamesonibett," followed, and perhaps associated, with Liparoceras Henleyi and Liparoceras BecheL Opliioceras raricostatxvm^ with its keel and septa, development and form, allying it closely with the Arietidœ, and indicating that its true position is at. the head of a series of this family, occupied geologically an earlier position in the " Raricostatusbett" of the lower Lias, than the two species which resemble it in the middle Lias. These are, undoubtedly, part of the Armß. capricornus of Oppel, and are there¬ fore found in the " Davoibett " of that formation. The planicostan abdomen which occurs occasionally in the young of raricostatum before the keel appears, leads to the conclusion, if we credit the hypothesis of evolution, that Microceras crescens and Mi¬ croceras arcigerens derived their peculiarities from the same source, and are either directly or iudirectly the descendants of this or some other commoo ancestor. I am disposed to credit the latter supposi¬ tion. The septa examined were those of young specimens, and in the case of the last named it will be noticed that the resemblance is remarkable in the external features of the shell as well as the septa- No one, however, I am confident, without having subjected thean to the closest scrutiny would suspect that they could be separated from Microceras laticosta^ with which they are also associated geologically. Again, this species is genetically connected with Microceras biferum on the one side and with Androgynoceras hybridum on the otlier. According to Quenstcdt, the former species is hardly separable in some of its varieties from Opliioceras raricostatum, but if the septa are examined closely they are found to difier, and the young arc dilferent. 1 Microceras laticostu here includes also the Amm. capricornus of Oppel and the two species alluded to in the note above, as M. maculatum and M. siTVuosUiTi. Hyatt.] 26 [October 6, The superior lateral lobes of Microcercts hiferum always, even in the young, seem to possess a median, minor cell which is absent m 0. raricostatum. The latter species is much the largest, and the adult septa differ widely. No genetic connection is traceable in their de¬ velopment except in very general terms. On the other hand, the affinities of 71/. hiferum in all respects point them out as degraded and dwarfed descendants of Microderoceras which precedes them, also, in time. There are other forms, however, which render these questions still more puzzling. A series of single spined or armatus-like species be¬ gins with Deroceras planicosiatum^ Dudressieri and Deroceras zipldus in the "Obtususbett," and is continued by Deroceras coifusum in the Haricostatusbett." The development of Deroceras armaiuin does not join it directly with any of these species, and since it occurs only in the lower bed of the middle Lias it need not be considered in this connection. Deroceras Dudressieri has the planicostan abdomen in the young, but in tlic adult possesses the abdomen of Microderoceras Birchil^ and in fact diflers from that species at this stage principally by the ab¬ sence of the inner line of spines; the septa are very similar in both. Deroceras zipldus differs more widely from Microderoceras Blrchii than Deroceras Dudressieri^ but in features which it is not important to discuss here. Then we have Deroceras planicosta^ wdiich never parts with the typical planicostan abdomen, though in the adult it acquires a single row of spines, as in Deroceras Dudressieri\ and lastly, Deroceras coofusum {Amm. Lolibergensis Emerson), which differs somewhat from/). in the septa, but more in the slighter form of the whorl. If now, we examine closely the development of the septa in Micro¬ deroceras Blrchii^ we find that it equally resembles the development of the septa in all of the members of the two series just described, which exhibit the planicostan abdomen largely in their growth. The septa of Microderoceras Blrchii on the first quarter of the third whorl acquires three minor cells, and the superior lateral lobes become divided, first bv the rise of minor cells from the sides of the superior lateral cells. During the same stage a very minute crenu- lation becomes developed from the side of the inferior lateral cell; this, however, does not increase as fast as its opposing cell, which eventually reaches a very large size, equally dividing the superior lateral lobes. 1870.1 27 [Hyatt In Deroceras Dudressieri this process is repeated at about the same period, but the dividing cell does not reach a similar prominence, nor do the septa in general terms become quite as complicated as those of the adult Microderoceras Blrchii until a much later period. Thus, ■while the lobes and cells of the former have become almost as com¬ plicated as in the adult, on the last quarter of the fourth whorl, those of Deroceras Dudressieri are a full volution later in reaching the same stage, and are never so deeply cut or foliaceous even in the adult as in the adult of Microderoceras BircML The first stage in the development of the latter corresponds to one which occurs in a precisely similar manner in Deroceras planicosta^ but not until that species nearly reaches the completion of its fourth whorl. In Deroceras confusum there is no constancy in the de¬ velopment of the minor cells. Two opposing cells may be brought out unequally, as in the young Birchii^ or symmetrically, or only one, invariably that from the side of the superior lateral cell. In other words, the adults have all the modes of division found in the différent stages of growth of Birchii^ according to the stage at which arrest of development has occurred. In neither Deroceras planicosta or Deroceras confusum do the septa reach a stage of complication com¬ parable with any bat the youthful stages of Deroceras Dudressieri and Deroceras Bircldi. D. ziphius was not examined, but the septa prob¬ ably accord with the growth of the external ornaments and pilm which place it near D. Dudressieri. The condition of D. Dudressieri and D. ziphius in the adult stage corresponds in their single external line of spines and rounded abdomen to the early stage of M. Blrchii, before the internal line of spines is brought out; that of the adults of Deroceras planicosta and D. confusum to the young of these two species when the spines are developed, and the abdomens still have the plaTiicostan folds. This characteristic, it will be remembered, occurs also in some specimens of Microderoceras Birchii, but is only faintly expressed ; in Deroceras Dudressieri and Deroceras ziphius it is constantly expressed in the young, to a later period in the former than in the latter, and is of constant adult value in Deroceras plani¬ costa and Deroceras confusum. The inference seems to be unavoid¬ able that the species of this series, which occur later in time and are all smaller than Microderoceras Birchii, are really dwarfed and de¬ graded descendants of this comprehensive species. Considering the septa in the next series, we have first Microce- TCLS hiferum. The superior lateral lobes in this species constantly Hyatt.] 28 [October^ divided equally, as in the adult of Microderoceras BîrcTiîî; the supe¬ rior lateral cells are divided into two unequal portions by a large minor cell, and are very similar in outline to the young of Deroceras Dudressieri on the fifth whorl, and to the young of Microderoceras Birchii at an earlier period, while the cells are broader and less deeply cut than they were observed to be upon the latter part of the fourth whorl. The young of ÄL laticosta are precisely similar to the young and adult of Microceras biferum^ but the septa bring out equally the opposing median cells, and the superior lateral lobes thus become unequally divided. In the adults they reach a state of com- ' plication comparable to those of Microderoceras Birchii and Deroceras Dudressieri. With Microceras laticosta are associated the strongly reversionary species, which only need a keel to be classified with the Arietidae. This is especially the case with Microceras arcigerens, whose septa, in one specimen, are remarkably similar in proportions and outline to those of Asieroceras ohtusum^ and what is still more remarkable in this same specimen, a slightly raised siphonal line is plainly apparent between the prominent planicostan folds. ^ In Androgynoceras hjhridum an equally complicated state of the septa is reached at an early stage, and still earlier in the succeeding species of Liparoceras. Microceras hlferum is of small size, about an inch in diameter, and at the latest stage assumes a double row of spines, or is smoother; the pilse in all cases closely simulating those of the adult Microderoceras Birchii at this period. Thus it may be said to play the same part that Deroceras Dudressieri does in the armatoid or single spined series in its external characteristics and form, while in its septa it corresponds to Deroceras planicosta. In the same way Deroceras laticosta may or may not have the double row of spines, but never has a single rowand never in the adult returns to the rounded O ' abdomen and peculiar pilae and ornaments of Microderoceras BirchiL Androgynoceras^ however, does return to this condition in the adult, but at the same time another tendency is developed both in 1A close comparison with Zieten Amm. Tumeri, which I regard as a variety of Asteroceras obtusum, shows, however, that a discrepancy exists in the proportions of the abdominal lobe and in the remaining general characteristics of form, which do not permit any attempt to trace a direct genetic connection. ^Microceras blferum occasionally has a broad projection on the pilae which might be mistaken for a single spine, whereas it is really formed by the coalescing or arrested development of two rows. 1870.] 29 [Hyatt. the form and septa. One is a greater degree of involution, the outer whorls as they grow, beginning to spread laterally over the sides of the inner whorls, and the septa keeping pace with the increased breadth of the sides, adding to the number of the auxiliary or inner lobes and cells. This higher degree of complication is carried to its greatest development in Liparoceras ; which, however, in its highest species, Liparoceras Bechei^ returns wholly to the mode of growth originally observed in Microderoceras Birckii, It proceeds directly from the young, smooth, round-abdomened stage, to produce the double row of tubercles, without the interpolation of the planicostan characteristics. It may be possible that the planicostan stage occurs in some individuals, but this would only complete the parallel with Blrchil which sometimes faintly expresses this reversionary feature. The conclusion with reference to this series appears to be, that its members are also at first degraded descendants of BircJiiij but instead of steadily decreasing in size and ceasing to exist, they first decrease and then speedily increase in size again, adding new elements of complication to the mode of involution, and increasing the number of the lobes and cells. All my attempts to trace a direct connection with those members of the Psiloceratidœ and Arietidae, which approximate to these series, have signally failed. The planicostan abdomen and the similar septa and forms which are found in the adults of Psiloceras, Caloceras,^ and Ophtoceras^ and in Microceras and Deroceras can be viewed merely as reversions, indicating, as in the difierent breeds of pigeons, only a common ancestry. It should be observed also, that where reversion is apparently piled upon reversion, as for instance, in the return of the Birchean characteristics in Androgynoceras^ after an interval caused by the pre¬ potent development of the planicostan abdomen, and an interval of time also, that this is not a reversion at all. It is, in fact, the re¬ sumption of a normal tendency beneficial to the race, which for a time has been entirely suppressed by the prepotent influence of a true reversionary feature. This can be doubly proved. In the Deroceran series, where no tendency to increased complication or size is observable, the race becomes enfeebled and dies out almost immediately. In the Microce- ran series, where a constant effort is observable to retain the double i A new genus, of which Caloceras torus and tortilis are types. Hyatt.] 30 [October 6, row of spines, to complicate the septa and increase the size, the law of acceleration is brought into full play, and overcoming the tend¬ ency of the species to be arrested in development both of size and characteristics, counteracts this tendency and reproduces the usual or natural succession of forms and characteristics. This may be substantiated in any series of Ammonites. By com¬ paring the lower forms with the higher of the same series it will be found that in most instances, when the series is complete, the spe¬ cies, as in Androgynoceras and Liparoceras, increase the extent of the involution and the number of lobes. This is precisely what occurs in the Arictidm, which are even more successful in suppress¬ ing the reversionary planicostan tendencies than the Microceran series. In this family the higher forms, Asteroceras stellare^ Asteroceras ac- celeratum ^ and others, are much more involuted than any of the lower forms, and this is still more strongly expressed in their descendants, the Amaltheoidaj and Ilildoceratidm of the middle and upper Lias. It may be objected that Mlcroceras hiferum is a young form of which we do not yet know the adult. Its size, the limited number of the whorls and the likeness of the septa, in the full grown speci¬ men, to the young oí Derocerns Dudressleri uná. Mlcroderoceras Birchii might be considered as proof of this supposition. The development is just intermediate between that of laticosta and Bircliii'^ any larger forms could therefore only intensify this relation. Besides the negative evidence, however, that no large specimens have ever been found, there is something positive. The possession of prominent tubercles makes it probable that quite an advanced stage of life is reached, since at a corresponding age in laticm^ia no spines are yet developed. Similar doubts with regard to the size of planicosta and corfusum in theDeroceran series are answered with more difficulty. The grad¬ ual decrease in size which the series makes from Microderoceras Bircldi through Z). Dudressierif Z). ztph'ius and D. planicosia to D. co)ifiisum, in all the dimensions of its whorls, when the full sized shells are considered, and the fact that these species, especially Z). planicostay have been very extensively collected, appear to make it probable that we now know the shells as they occurred in the localities and » strata in which the}' are found. That they may be dwarfed speci- 1 Kew species, which has the abdomen like Aster obtusus, but is more involute than any other species of Arietes. 1870.] 31 [n7att. mens which did not develop beyond periods corresponding to the younger st Lolibergensis Emerson, Die Liasmulde von Markoldendorf, p. 61, pi. 3, fig. 3. In this species the first three and a half whorls are smooth and flattened ventrally, the sides bulging as in 0. raricostalum. This resemblance is still further increased by the development of the pilm. On the latter part of the fourth, fifth and sixth whorls the resemblance to raricostatum is very close, or rather to the earlier stages of that species before the keel appears. On the sixth whorl the tubercles begin to appear and the form changes to a more laterally compressed and thinner "whorl, and the tuberculated pilœ cross the abdomen as in the iyficdX planicosta. The septa on the fifth whorl are quite like those of raricostatum in their outlines, though the inferior auxiliary lobes and cells slope inwardly and posteriorly. All the shells examined were small, hardly more than an inch in diameter. The developmental resem¬ blance to 0. raricostatum does not extend to the sei^ta. These have a close similaj'ity to those of Caloceras torus, differing however in one H7att*] 40 [October 6> essential point,—the presence of median minor cells which equally divide the superior lateral lobes. This characteristic, though it may be absent in many specimens, is so constant that it prevents the direct connection of the young of this species with the young of C. torus or 0. raricostatum^ which it otherwise so closely imitates. The resemblance to the young of Microderoceras Bircliii is not so close, however, in external features, though the septa are very closely allied. The Turrillite variety is quite common in this species, whereas it is not so common in the true Amm, planicosta Sow. One of Quenstedt's figures of this species represents a young shell decidedly keeled. This, I think, cannot be of the same species, and his accompanying descriptions do not justify its associations with his Amm. BronniL TTae variations in the lobes are excessive. The median cells of the superior lateral lobes are usually largely developed, often, though not invariably, retaining the youthful or one-sided aspect which they have in the young of Dereceras planicosta and M. BircJiiL In some specimens, however, they are very small, and the lobes are unequally divided by two very minute minor cells. These lobes, in other words, may be equally divided, or have all the gradations from this to a state of unequal division. The same lobes are either longer, equal to, or shorter than the abdominal lobe, but seem invariably to greatly ex¬ ceed the inferior lateral lobes. Deroceras densinodum. Amm, armaius densinodus Quenstedt, der Jura, p. 105, pi. 13, figs. 9, 10. D. densinodum Hyatt, Bull. Mus. Comp. Zool., no. 5, p. 84. This species does not apparently occur in the middle Lias as stated in my paper in the Bulletin of the Museum of Comparative Zoology, but only in the lower Lias. The mistake resulted from an erroneous reading of the label on the specimen. It may be only a variety of D, confusum^ but the young differ somewhat, the abdomen is nar¬ rower and the septa are invisible in the single specimen of D, densi¬ nodum in the possession of the Museum. According to Quenstedt's figure this species is really an armaius in which the young is pilated or ribbed at a very early period, instead of being smooth as in armaius proper. Having only the young it is impossible to say much about the affinities of the shell ; it is, how¬ ever, evidently a member of the planicostan group or series of Deroceras, 1870.] 41 [Hyatt. ANDR0GYN0CERA8. From the specific descriptions it will be seen that we have here two groups or series, both developing from the first variety of one species, Androgynoceras hybrid.um. From this we have the series in which acceleration of development produces the flattened abdomens and broad whorls of the second variety of A. hybridum^ and of Liparoceras indecisum^ the more elevated, though still broad whorled adult of L. Ilenleyij with young just like the adults of A. hyhridum^ second variety, and finally the high whorled L. BeclieL The amount of involution is just proportionate to this progress, reaching to the first line of tubercles in the first three forms, to the second only in the adults of the fourth, and to the second in both the young and adults of the fifth. The second offshoot or series contains only one species, A, appressum, which is highly accelerated when compared with A. hybridum. This has remarkably fiattened sides and the connection with the other is shown by the development of the young. Whether this ought to be set aside as a distinct genus, or not, will depend upon the discovery of other descendants. Androgynoeeras hybridum. Amm. hybrida D'Orb., Terr. Jurass. Ceph., p. 285, pi. 85. And. hybridum Hyatt, Bull. Mus. Comp. ZooL, no. 5, p. 83. For six volutions the shell is apparently inseparable from certain forms of M, laticosta. Upon the seventh whorl, instead of continuing the same degree of increase in size, a more rapid enlargement takes place, the lateral pilae become less prominent and more crowded, sometimes coalcscins near the umbilicus. The inner tubercles in the latter case, also, either partially or entirely coalesce. The abdomen in the meantime has become more prominent, less furrowed, and more rounded, and the sides converge outwardly. The abdominal pilae split up each into several minor ridges on the latter part of the seventh whorl, reducing these furrows to a minimum. A variety of this species from Schippenstadt and Semur completes the same stages of growth as have been described in A. hybri¬ dum^ a full volution earlier. It has at the end of the sixth volution a whorl nearly as large and of the same form, but much broader in pro¬ portion to the length than in the first. The pila? begin to take upon themselves similar characteristics. I am unable to state whether this or some intermediate form between this and the first variety is the Hyatt.] 42 [October 5, one described by D'Orbigny ; or with any certainty, whether the first variety is a distinct species, though it seems to be such. Androgynoceras appressum. And. appressum Hyatt, Bulb ]\Ius. Comp. Zool., no. 5, p. 83. For perhaps the first four or five volutions the shell is similar to the young of thinner varieties of M, latlcosta. The extent of envelopment, also, is slight ; soon however, on the last of the fifth or first of the sixth the planicostan pilse split into several ridges united at either end by tubercles. The abdomen at this period, the inclina¬ tions of the sides, etc., closely resemble the characteristics of the adult of the first variety of A. hyhridum. LIPAÄOCERAS, Liparoceras indecisum. Lip. indecisum Hyatt, Bull. Mus. Comp. Zool., no. 5, p. 8. This is evidently a very much more accelerated form than even the second variety of A. hyhridum. It still preserves, however, the form of the adult of this species. The young, if I am correct in referring a young specimen from Bautenberg to this species, has plan¬ icostan pilas certainly until after the completion of the fourth, and probably until near the end of the fifth volution. Liparoceras Henleyi. Amm. Henleyi Sow., Älin. Conch., vol. ii, p. 161, pi. 172. NauU striatus Bein, Naut. et argo., p. 85, pi. 8, figs. 65, 66. L. Henleyi Hyatt, Bull. Mus. Comp. Zool., no. 5. p. 8-t. This species differs from the last in not repeating the planicostan abdomen at all, unless upon a volution preceding the last quarter of the third. This seems improbable, though it may occur in some speci¬ mens. At this period in the specimens examined all the adult char¬ acteristics were well developed, and it only remained for them to increase in size. The form of the shell is also precisely that of the adult A. hyhridum^ second variety, or at least of that shell at the end of the sixth volution. The L. Henleyi differs from L. Bechei in having larger and more prominent tubercles, a broader whorl in proportion to the height, and in the slighter involution of the young. For the first four whorls the involution does not reach the internal line of tubercles, leaving a larger umbilicus than in L. Bechei. In one specimen from Lyme Regis the only distinction from Bechei consists in this single 187).] 43 [Hyatt characteristic. Usually, however, the angular, prominent, ribbed tubercles of the younger whorls at once show great differences. While L. Henleyi thus seems to show variations advancing towards L. Bcchei^ the last never has a variety like L, HenleyL Liparoceras Bechei. Amm. Bechei Sow., Min. Conch., v. iii, p. 143, pi, 280. " Ziet. Verst. Wart., p. 37, pl. 28, fig. 4. Lip. Bechei Hyatt, Bull. Mus. Comp. ZooL, no. 5, pi. 84, Fold-like lines of growth are prominent in the young, which are otherwise smooth and rounded. These increase in number and sharpness until they become true pilae. Two lines of tubercles are introduced, also, in the young, but appear either quite late or comparatively early on the third whorl. The septa on this volution precisely resemble the septa of the nearly adult M. laticosta. They have very broad abdominal and superior lateral lobes; the latter unequally divided. The siphonal cells are very large. In one specimen a very decided resemblance to A. sternalis is produced by the angulation of the abdomen and the unusual development, for so young specimens, of the lateral pilae with their tubercles on the last quarter of the third whorl. ^ 4- I'rom the rnx-t'tMlings of the Knston Society of Natural History, \ ol. XVIIl pp. ;{8()-i01, June 7, 1S70. GKN KTlC KELA ri( )N-S OF STEPHANOCERAS By A. 1JYATJ\ ' "iN^.T.N, LJL \ From the Proceedings of the Boston :"r.<'icty i>£ I«ialiirul History, Vol. XVIII, pp. 3'')'i—1ÍIÜ, June 7, Jb7(>. GEIV E11C E ELATIONS OF STEPHANOCERAS. BY A. HYATT. The group which forms the subject of the present paper was first described by Waagen as part of his genus Stephanoceras, it being with Dactylioceras coimnnKP and its allies, united as the '*sub-group a" under this name.^ Cœloceras Petto^ was left by the same author under the title of .<.Ho-oceras, though the similarities of the latter to ' r> Steph, Humplirteslanuni were fully recognized by Waagen in his sub¬ sequent paper.2 In this pap»;r, also, he restricted the use of the name Stephanoceras; and two groups, whidi had appeared as sub¬ genera of Stephanoceras in his first pa[)er, were elevated to the rank of full genera, under the mimes of Kosmoceras and Perisphinctes. The preservation of zoological nomenclature in an available form demands above all things that names shall not be uselessly multiplied, ^ Die Formenreihe d. Arnm. suhradiatus. Benecke's Geog. Pal. Beiträge, Vol. 2, p. 248. 2 Ueb. d. Ansatzstelle d. Haftsmuskeln b. Naut. und Amm.; Paleontographica, Vol. 17, 5, p. 215. 9 and for that reason the hiw of priority has been universally recog¬ nized and mercilessly applied, Waagen, and all other Germa» Paleontologists who have tpioted his names, have disregarded this law in a wholesale nianirer. The only reason for this conduct, and the most charitable one which can he given, is, that they considered the new names proposed hy Prof. Agassiz and myself as untenable^ and unworthy of their adoption. This reason, although perhaps suffi¬ cient to themselves and their followers, is no justification for a viola¬ tion of the rights of prrorrty. The laws of nomenclature do not permit them to describe the same family groups as new genera with new names. New views of the relations of well known species can¬ not be represented hy¿e^^ generic names because the grouping happens to include a hali^ozen more of the previously described genera. What a fearful maze of di^iculties this process would lead to if generally adopted! Every man, or set of men, would of course have the same privilege. For example, let us suppose that in my own recent paper on the " Genetic Relations of the Angulatidae," in the Pri>ceed.of the Best. Soc. Nat. History, Vol. xvir, May, 1874, I liad originated a new name for the genus ^Egoeeras of Waagen, O O S ' because his generic characteristics are of no value for the distinction of grouj)s of generic significance. The genus jEgoceras, according to AVaagen, contains forms as widely separated as Psiloceras plan- orhis^ belonging to the Arietidse, JEgocera^ a7\gulatm, one of the Angulatidae, Androgynoceras Ilenleyi^ one of the Liparoceratidae, and Cœloceras Petloí^ of the Dactyloida?. According to their development, mode of occurrence in time, and all their adult characteristics, except perhaps " the undivided, horny character of the Aptychus," these forms are perfectly distinct from each other. The Psiloceras becomes the parent form of the Arietidae in the Lias, the jpgocei-as angulaium of another distinct series differing wholly in development and form in the same formation. Both of these are probably traceable to a common ancestor in the Trias, according to AVaagen and Mojsisovics,^ and therefore it may perhaps be considered that it is legitimate to join them, but what can be said with regard to the remaining forms? Androgynoceras Henleyi is di¬ rectly traceable to Deroceras Dudra^Aieri^ the affinities of which can¬ not be settled with our present knowledge conclusively; but what evidence there is, however, in the development of the young shows iSee also my paper on the " Genetic Relations of the Angulatid«," in these Proceedings, Vol. xvii, p. 15. KRost decidedly, as be anticipated from tke adult characteris¬ tics, that the ancestral fonns are to be sought in the Lytoceras and allied groups, not in Psiloceran forms of the Trias, C^^loceras Pettas is equally of uncertain derivation, though its affmities i-n every re¬ spect show also that it belongs to the Dudressieri se-ries. All of these forms are Included under the name Stephanoceras, and thus two great groups of Ammonites, the round abdomened and the keeled groups, with distinct systems of development and uncer¬ tain derivation are made to appear as one genetically connected series. This, however, would not justify the total suppression of the name JEgoceras and the substitution of another for the more limited group, to which it -can be properly applied. Scientific courtesy, as well as the strict law of custom, forbids such a course, tliough here, as in the Arietidae, I must consider the name as used by Waagen utterly devoid of zoological meaning, Tl>e structiii-e of the Aptychus has, no doubt, some meaning, but it alone certainly cannot unite PdL planorhÍ!^^ jc^goc, angula¡uí<^ Cœlocei'as Petios^ j\fici'^(l€}\^eras etc., into one genus, because as Waagen himself points out, it has the same structure in two other groups, Arietites and Amaltheus, de¬ scribed by him as distinct genera. If he had joined all these into one group and distinguished thein by the Aptychus, it would have been more consistent and less objectionable; this characteristic would have at any rate applied to them all, I allude particularly to this fact because the other characteristics given by Waagen are not applicable to such large groups. Thus in the lower forms of the Arietidae (that is to say, my genera Psiloceras, Caloceras and Vermiceras, including the planorhh^ raricoslatxis and Conyhearl series), the length of the living chamber, one of Waageu's distinctive characteristics, is generally over one volution. In the genus Arnioceras, the falcarles series, its length is generally less than one volution, from one half to nearly a full volution. In Coroniceras, from one-half to one. In Asleroceras ohtuswn the length is from one- half to five-eighths of a volution in large specimens, in Asteroceras Brook'd about three-fourths. In Aga^dceras lœvîgatus^ five-eighths to three-fourths of a volution, in Agassicei'os Scipionianus, about three-fourths. Thus in all the higher genera of the same family it is less than one volution, and so variable that it cannot be very use¬ fully employed, even as a specific characteristic in some species, such Aster, ohtusum. The outline of the mouth has been long used to designate sub- o ^ groups among the Ammonites. This characteristic, like all others, is i m 0 of différent values in diíTerent and the attempt to use it with the same me-iniuL; in evt ry results, as in all otlier cases, in coufiisiori. Thus in Waa'^eu's dia;:;iioses of the j^enera Stephanoce- ras, Perisphinetes and Kosmoceras, we find that they are all three described as liavinfj; simple (entire) mi)ntl)-openines or ears." In each . In one school the effort is being ]>er])etually mad.eto di^co\er s )me set of characteristics by which animals mav be di^tingni^hi'd one Irom another. Every new organ, or indication of ^uch, wlien dNeovered, is af)plied at once to the definition of groups, as if this was the great objt»ct of all classifi¬ cation. The distinction of groups from each other doulitless repre¬ sents to a certain extent our knowledge of their organization, but only in proportion to the number of the parts or characteristics which may be eiupUyyed in clas>iricall'm. Conscpiently arbitrary classifica¬ tions based on single (diaraeteristies are the most imi)eriect,since they necessarilv leave out of coiisideratloii the numberless allinities of the « groups, and all the minor points of diiferenee which here and there appear. m b In the other school, a zoologist or paleontologist makes greater allowance for the variabilitv of ofiianic bo'lies, becomes distrustful of all single characteristics, or combination of single characteristics, and endeavors to combine all possible sources of information in his defi¬ nition of groups. The former naturally tends to the formation of large generic groups, those which can be approximately distinguished by some salient structural characteristic, and the latter to the division of these large groups into many minor ones, in order to show the nat¬ ural affinities and derivation of animal forms. The former leads to the artificial method of classification which has always, witliout the slightest reason, been claimed to be the more useful, antl the latter to the approximately natural method. The difïer- ences are most prominently ])resentcd in one, and in the other these are considered of no more importance than any other class of charac¬ teristics. The first is certainly the most imperfect and conventional; and why its defecUs, wliiidi are openly confessed, should be regarded as recommendations for its a* < fcí ÍA ATHLirTA-nEI> 8. subleve Anceps-beü S. macro- cephalum S. dimorphnm S. Ilerveyi S. píatysío- inum Macrooephalus BED S. subleve >5 O H <1 g tí O h H < tí S. microS' tomtm S.Bombur LAGEÍÍA-BED S. plauulum S.eoToriatum Digona-bed l g M tí o o tí tí O tí S. plicatissimum Pabkinsoni-bed S. coronatum S. Deslongchampsii S. microsío- mum rar.Ymir S. contractum S. coronatum HüMPnniESiANüS- BKD S. iilagdeni S.Brongn- iartii S. Gervilii S. lingu- iferum S. Huniphriesiauum S. Brochií S. contractum S. Sauzei S. llraikcnrídgii S. liayleanum S. subcoronatum S. nodosum early as lí^4r) he ado^jted the mode of work which is fast becoming univeisal, that of uiiiiin.ü in the same genetic series all forms, however dissimilar in aspect, which can be traced into each other, by means of the vonnii and of the adnlt characteristics. Stephanoceras 1 Waagen (Fars). The earliest observed form of this genus is the Stepk. nodosum = Humphriesianus nodosus Qnenst., which occurs in the Humphriesianus- bed. This variety or species, wlnchever the taste of the reader prefers, has the ribs more prominent and more widely separated than in Humphriesianuyn^ the umbilicus larger, and the whorls increasing more slowlv in size by growth; this renders the shell altogether more discoidal in aspect. The varieties, however, show a shading of the characteristics in three différent direction«. One way leads to Steph* Bayleanum^ and another to Sft^ph, Hnmpliriesidnnm, and still another to subcoro)intu)if. Towards an ii/n a retrogi-ade series of changes produces forms more and more discoidal, with whorls increasing more and more slowly in size by growth, until in the typical Bayleanum a very distinct species aj>pears, as figured by D'Orbigny, and discussed by Oppel. It occurs contemporaneously with nodosam, and also later in the up])er j)art of the Humphriesianiis-bed. In a similar way, by following the indications of the gradually changing varieties we are led to the stoutest, most involute and co ' narrow-umbilicated forms of the typical Sfeph. Humphriesianum. In these the abdomen is also more elevated and rounder, the ribs are finer and more numerous, and the sutures distinct. Steph. suhcoronatum, as pointed out by Opj)el and Quenstedt, is one of the transition forms of Humphriesianwn, but it has a wider significance when carefully studied in all its varieties. It becomes identical with Amryi. Deslong- champsii when the ribs are curved and prominently tuberculatedy and the abdomen soinewbat elevated, though still very broad. The abdomen becomes in some specimens still more elevated, the umbili¬ cus narrower than in the Deslongchimpsii, tbe umbilical shoulder of the whorl more abrupt, the umbilicus deeper, the abdominal ribs par- 1 This name, as has been pointed out to me by E. B. Tawney, Esq., of Bristol has been already occupied by Khrenberg for a genus of Rotatoria, but the termina¬ tion adopted was spelled with an " o " instead of an " a." Stephanoceros instead of Stephanoceras, and this seems to me quite sufficient under the circumstances to justify its retention. 9 ticularly fine and numerous, the lateral ribs like those of Steph. Un- guiferum or those of Steph. Deslungcharnpm. Tliese are apparently identical with the pVicatissiinum of Quenstedt. Both of these forms, Steph pUcadiisimum and Steph. Desloitgchampsii.! are found in the Parkinsoni-bed. Some of the varieties of Steph. suhcoronatum are nearly identical ■with Steph. nodoiiam^ and some of them resemble closely the smaller specimens of Steph. corottatmn or Bhtgdeni, The forms from Dundry, and also those alluded to in Quenstedt^s descriptions of Iluniphriesi' anuSy as allied to Amm. Brocchii Sow., show a close scries of transi¬ tions from the finer ribbed s})ecimens with open umbilici anhrifsianuin. I shall have frcijuent occasion to speak of, and as its resemblance is general rather than special, I shall sj)eak of it usually as the Pettos-like form, in allu>ion to its ancestral derivation. S. Blagdeni may be briefiy described as a huge form of a young suh-coronatum the broad abdomened variety in some of its forms; in others, however, the abdomen becomes elevated, and no line can be drawn between these and the succeeding, or true Sfe/di. coronatum series. The peculiar broad abdomened f(>rms which began to a]>[)ear in varieties of Stepit, suhcoronatum are in Blagdeni^ the predominate ones, and represent the species. The young changes hut slightly by 10 growth, except in size, and the Pettos-like form is retained through¬ out life, except in those varieties which approximate to coronatum^ or more strictly speaking, except in the round abdomeiied varieties' which approximate to the predominant round alxlomened forms of Sleph. coronatum. These last do not alter the peculiar coarse charac¬ ter of the lateral ribs and tubendes of Bla(jdenî, but simply elevate the abdomen an I increase in size faster by growth than the normal varieties, so that the umbilici become narrower, and the sides of the whorls more abrupt. These are often called Amm. Banksii Sow., but may be distinguished by the young which have the fiat abdomen of the true Blaydeni until a late period of growth, while the true Banksii has young with a inoi'e elevated abdomen and larger tuber- » O O des. The Pettos-like form of Blagdmi and its peculiar ribs are more or less represented in all the young forms of the true Steph. coro¬ natum» Sometimes specimens retain this even to an exaggerated degree, growing up to the adult condition with the sides so sharp, umbilici so deep, and abdomens so flat, that they appear as new specific forms, until the connection is traced between them and the normal forms. These are, as in the case of the similar representative forms founil| in Sfeph. subcoronatum» generally rather small; such is the variety known as the anceps-oniati of Quenstedt, and other scat¬ tered varieties intermediate between this and the true broad abdo- mened coronatum forms. Both Sitph. subcoronatum and Blaydeni occur associated in the Humj)hriesianus-bed, and Sti^ph, coronatum later in tiie P>ion of an evidently inherited tend¬ ency is acc()m])niied hy a correllative tendency to the suj)pression or absorption of the tubercles ami ribs. These changes are retrograde in so far as they produce a tbrin smaller and less ornate than the preeetion very fully in his diagnosis of the convolutus group, p. 578 of Der Jura, am) it is also my impre>sii>n, derived from careful examination of closely allied forms, that even such apparently i'oron(ituinA.\k.Q, forms as tiie Graveai- anum^ figured by D'Orhigny, pi. 219 of Terr. Jurassique, will be found to belong to the roiicnlutnui or planulafu/n group, and that the true coronafi have no representatives in the White Jura. The extraordinary form, Steph. suhl(ece, to which we now come, presents in its adult condition so close a resemblance to the Amm, GoUnlhns that Quenstedt is evidently in Der Jura " doubtful of its true afïiiiities, though he hatl j)rcviously, in Die Cephaloden," re¬ ferred it to the coroiiafu/n group. The development, however, shows none of the peculiar variations observable in the Amm. Goliathus groiq>, and the young in some specimens retain the coronalum form and characteristics until a late stage of growth. During old age the whorl contracts as it does in Humphriesianum. The form and char¬ acteristics of the young appear to indicate a derivation from some voronntum form, like that found in the Parkinsoni-bed, Museum of Stuttgart Collection. Another characteristic which seems to separate it from the GoVuithui^ series is the general tendency of most of the forms to become iiuooth on the abdomen, at a stage when Goliathus is furnished with prominent ribs. ííotwitlistanding these facts, how¬ ever, whenever the adult forms come under observation, a similarity becomes a[)parent which it is at present im¡)ossible to explain. The series which can be followed from Steph. contraclwn to Brocchii, and its allied forms, is perha]>s the must complete of all others, the ' With this coinjiare the old coi^onafum described by D'Orbigny, referred to above. 13 lines drawn between the different species being so slight that they vary with every locality. Confracium can only be separated from subcoronafum by the fineness of the ribs on the abdomen, and in the adult by the aspect of the sides. The connection with subcoronafum is largelymade through the young, which are indistinguishable from the young of that species in some specimens. The Ilerveyi-like, or macrocephalum-like forms of contractum occur¬ ring in the Parkinsoni-bed, have finer ribs than llervefji^ but it is probable that they vary greatly in this respect. The young of Sfepli, HerveyizxQ in some varieties tuberculated, but acquire the aspect and characteristics of the adult of Sleph. contractum^ including the fine abdominal ribs, as soon as they lose their tubercles. Others which have no tubercles acquire this aspect at still earlier age, and these lead into Stepli, macrocephalum, in which the young are invariably smooth, or not tuberculated. In Steph. macrocephahim we find a series of forms, which become gradually more and more compressed laterally, until they present a very narrow abdomen and whorls of extraordinary breadth. The abdomen, however, does not become sharp, even in extreme varieties. Throughotit this series, as a rule, only the oldest specimens become smooth on the latter part of the living chamber, showing that this is an old age characteristic. The growth maintains the same ratio of O O increase in the size of the animal throughout life, and the whorl therefore never becomes contracted even in extreme old age. There is, however, here, as in the compressed forms of other series, a notic- able decrease in the size of the species or varieties as a whole. The laterally compressed forms are usually mucli smaller than the broad abdomened forms, a fact in direct accordance with the idea that they are the senile descendants of the broader forms. The mouths of this series, like those of all species previously nphr¿f-,si]Lr¡<'siuiins pllcafus Quenst., Der Jura, p. 398. A/n7n. Humphrlrslanus D'Orb., Terr. Jurass., pi. 134-135. The typical forms are found in the IMiddle Brown Jura and in the Mus. Stutt. Coll., with the first of the true coronatum forms. The varieties appear to have two ])rinci{)al tendencies, one which leads into forms similar to HumphrltAanus plicafissimus Quenst., and occurs in the upper part of the same formation and one which approximates to the Amm. suhcorunatus Oppel. One fine specimen of this form showed an incomplete living ch.imber at the diameter of 156 mm., about half a volution in length. This was smaller in every way than the adjoining whorls, but no signs of old age were visible. The finest suites of this species occur at the Bristol Mu¬ seum and in D'Orbigny's collection. One specimen in the latter shows an extremely long and complete living chamber, occupying one and one quarter volutions. The entire diameter of the s})ecimen was 210 mm. The involution of the whorls was noticeably dccreasins: at about 80 mm., and continued steadily to decrease, accompanied by a corresponding diminution in the size of the whorl until the difference in size and form at the mouth became very marked. This specimen exhibited an extreme variation, and should be more exactly, perha[)s, associatcarison shows that he su[)})osed D'Orbig- ny's figure to represent a s])ecies closely allied to a/huslige- rus,** whereas it very accurati'lv shows the characteristics of a well ¥ ¥ known French foi-m which passes insensibly into coi-onafus^'^ and is found associated with the latter at Chatillon sur Saone in the Bath- formation of 0[)pel. The originals do not exist in D'Orbigny's col¬ lection, but young specimens show that their relations are probably correctly stated, as above. 31 FOURTH SERIES. Stephanooeras subleve. Amm. subleviii Sow., Min. Condi., pi. 54. Amm. modiolaris D'Orb., Terr. Juras., pl. 170. Amm. suhíeviíi Quenst., Die Cepli., pi. 14, f. 0. Amm. suhleoh Zieten, Yerst. Wurtt., pi. 28, fig. 5. The originals in Sowerby's collection prove the accuracy of Quenstedt's conclusions with regard to the identity of the English, French and German forms. D'Orbigny's collection possesses only a cast, but his figures are quite sufficient. Amm, suhlevh Zieten, which Quenstedt identifies with modiolark^ is represented by several specimens in the Upper Brown Jura, Ma- chrochilus-bed. Museum of Stuttgart. One of these is much thinner than the others, and shows a more discoidal young. The rest have very abrupt sides from an early period, and deep umbilicus, but not so deep as in D'Orbigny's figure. These show that the form is not developed as in Quenstediocera^ Leacliil, and others of the Goliathus group, to which tlie adult of the modlolare variety seems to be closely allied, but accorht ine back to the same conclusion that this was due entirely to the purely coronatum-like form of the young, which at a very early stage is not round and smooth as in Gervilid but more like suhcoronatnm or Blagdeni. This remarkable difference in the development confirms the contrast of structure between the mouth of the shell with its ear-like lappets, and the plain IIum{)hriesianus-like outline of that of Gerviiii The form also differs somewhat. The living chamber near the mouth becomes depressed from above, as in BralkenrldgH^ instead of contracting laterally, as in Gerviiii^ and all allied forms. There are several varieties, but the principal are those with open umbilici, in which the young retain the true L'oronatum form until a late stage of growth. These always seem to have prondnent tubercles at an early age, and are altogether more similar to Braikenridrjii than those with narrower umbilici. The last are more involute, have tlie tubercles later developed, the ribs finer, and the young in form and markings so similar to the young and adults of Gerviiii or Brongniar/H that they are often confounded. This is one of the few instances in which the history of the devel¬ opment and adult characteristics appears to be at variance with the geological record. Braikenridgii has only been found in the Ilum- phriesianus-bed, whereas Sauzei is habitually found in the lower part of the Ilumphriesianus-bed, the " Sowerbyii-bed." This, how¬ ever, is only a slight discrepancy which may arise from false identifi¬ cations, and I have therefore ventured to disregard it in the genea- ' Ö O logical table. S DOUBTFUL SERIES. Stephanoceras refractum. refractas Rein., Naut. et Argo, figs. 27-30. Amm. refractas D'Orb., Terr. Jurass., pi. 173. Quenst., Der Jura., p. 524, pi. 69. This bent and distorted form has young which can be compared only with the young of this series, and it is possible that a suffi¬ cient number of specimens would enable an observer to trace it directly to some one form. There is, perhaps, more resemblance to 41 microstomum in the young of the specimens which I have exam¬ ined, but the large ear-like lappets are very dissimilar character¬ istics. The abdominal channels are present in some specimens of microstomuifi, and in some of the other species of the group as a rare variation, so that their prominence in this species can not be consid¬ ered as absolutely conclusive against this view of the affinities. I Iiave failed entirely in finding any species of the Parkinsoni group to which the young might be compared. The development of the ears seem to decide in favor of its association with the Sauzei group, but the large rostrum between them is an entirely new organ, not shown in either Braikenridgii^ Sauzei or Unguiferum. In fact it has the most curious and unaccountable mingling of the characteristics of several groups, with certain prominent character¬ istics entirely peculiar to itself. Quenstedt quotes one form as found in the Parkinsoni-bed, and speaks of this in **Die Cephalopoden " as an undoubted " crippled " ParkinsonL I have failed to recognize this fact in his collection. My notes give me no hints on the subject, and I may have omitted seeing the specimens he refers to. Whether to connect this species with the Microstoma impressa Quenst, of the AVhite Jura or not, I cannot say. There appears to be a close affinity between the development of the young, and the abdominal furrow is well developed; but on the other hand such re¬ semblances might occur in simply representative species of distinct genetic series. The Amm. Schapliitoides Coynarti of the Oxford, fine specimens of which exist in the Prof. Mœsch's collection at Zurich, has an irregular form and the same furrow in the abdomen of the living chamber, but the mouth was not shown. Amm, Chap- uisi and Collinii Oppel of the White Jura of the same collection, are evidently closely allied to Amm, scaphitoides, but like that species resemble refractum only very remotely, and I think will be traced eventually to some form in the White Jura. 5 fV\Crv^_- íliL^v^JLi^ C^A^TNJU ^ C\7\/X_j£^ rULA/'VN.jCXjO'i^ uJyA-crvv_-A^rvJU'■ CiJLrvO^-, crR^^_^^:5UjU. O^JÛLxJuÀ- 5|^CVWV î) OwcXx^jC)kA_(XjL-. • *« / -N. ««./-«-S^ €V ^ NST ^ , ILL. [From the ProceediDgs of the Boston Society cf Natural History# Januaiy 17,1872.] The Non-Reversionarv Series of the lirarocfratid.e, ant) Rkimarks Upon the Series of the Allied Family Dactyloidaî. By A. Hyatt. In my last communication to this Society I endeavored to trace the reversionary series of the Liparoceratidœ, and necessarily left out of consideration all those species vrhich did not belong' to one or the other of the truly reversionary series. Of these last there is a very curious group, all the species of whichj at a young stage, resemble more or less closely the young of the original type, Microdoci-ras BircJtii, but betray by their de¬ velopment, adult characteristics and geological succession, a closer affinity to Derocej^as DitdrcssierL This species in the communication referred to "was shown to be the probable ancestor of a series of forms which include JJeroaras con- fnsum and heroceras d(ïisbiodiim. All of my observations, as well as those of Quenstedt, indicate that the true })osition of Deroctrm densinodum, is between Dei-oceras Dudres^ei'i and Dtroceras «ma- /wm, and its geological relation is similar. It is ibund in the Ra- ricostatusbed, just between the Obtususbed containing Beroceras: Dudressieri and Deroceras [danícosta^ and the Armatusbed of the Middle Lias. My material, however, does not permit me to com¬ pare the young of Deroceras arinaium and Deroceras denshkodum in order to ascertain whether the former is really more closely allied to, the latter than to Deroceras Dudres^teri, but Quenstedt sup¡)lies this deficiency in the proof. According to his figure, this species is really an armatus^ in which the young is pilated or ribbed at a very early period as in instead of being smooth as m armai us proper.^ The general tendency of the series, as shown by the adults and by the growth of individuals, is to produce the spines at wider inter-- vals, depress or obliterate the pilie, and by decreasing the breadth 2 of the abdomen render the whorl more c}dindrical. Deroceras Davœi candes this tendency still farther. Tlie young being smoother, the spines more s])arsely distril)uted, the fold-like pila3 almost absent in some specimens, and wholly superseded by the projecting lines of the transient mouths. Tliese having previously been merely subordinate pila?, have become primary and replaced the folds. Tt is not at all improbable that Deroceras allernum, from Plateau de Lazae, wliich aj)pears to be very closely allied to Deroce¬ ras Davœi, may prove to be a local variety of the latter. However this may be, the young are more like those of armatus than are those of Daroe Deroceras minatam is labelled by Dr. Krantz as a new species from the Upi)er Lias of the Plateau de Larzac. It differs in the form of the adult whorl and in the septa, though not in the young, from Deroceras alternum of the Middle Lias. The sides and abdomen are flattened instead of gibbous, as in the latter, and the shell is much flatter and thinner. The lobes are more deeply divided, and larger every way in Deroceras alfenium, and the cells narrower at the neck, and S2)reading more at the base than in Deroceras laiaatnra. In neither of these species are the s])ecimens probably fully matured, but yet in one specimen of Deroceras alter¬ num^ only a half inch in diameter, the changes are fully as extensive, and the septa as completely developed as in the adult of Davœi. The quadragonal form of the whorl and the tuberculated pibe occa¬ sion in the young Deroceras minatum, during the later stages of growth, a resemblance to Holandrei, which subsequently changes to a nearer resemblance to attnidaium. Another series begins with Deroceras muticum of the Lower Lias, which aj)pears in the Paricostatusbed, just above its nearest ally, Derimeras DudressierL This species is very much smaller than the last named, and diifers in other respects ; but it resembles it more closely than Deroceras armaiuni ; whether it compares in the same manner with Deroceras densinolum or not, I am unable to decide. Its adult characteristics, however, and development, while they ap¡)roximate to Deroceras arnwtum, do not permit of its association ill the same series. The absence of intermediate pila?, the con¬ stancy and closeness of the tuberculated pilae and the septa are very distinet in tliis species, and the huge Deroceras iiodoyi(jas of the Mid¬ dle Lias, which appears to form the second term in the same genetic series. Both of these series exhibit, so far as I can trace, no rever¬ sionary characteristics. "Whether the modifications may be consid- 3 ered progressive or not, remains to be determined. There is the same tendency as exists in the other series of the family previously described to suppress the fold-like pilie and prominent tubercles. It is noticeable that in the annatus series this takes the same direction as in the genetic series of the Dactyloidie, which begins with Duciy- liocenii^ commu'ie and ends with Dactyliacems Braunlatium. The tendency in both of tlu-'se groups is to obliterate the large tubercu- lated fold-like piUe, with the minor ridges or pila? gathered and join¬ ing in fascicles at the tubercles, and to substitute for this and the flattened abdomen a rounded whorl, single, sharj)1y defined, entire pila?. It must be remembered, also, that in neither of these series is there any increase of the involution in successive species. The umbilici are entirely open. The septa of Deroceras Bavcei are rather more ornate and complicated than those of Deroccrai^ annahtiny and the lobes and cells of the latter than those of Deroceras iJut/ressieri. A certain amount of progress may perhaps be claimed for this series, but the other series, that of Deroceras laaticunty is too incomplete. De¬ roceras muiicwn is certainly much smaller than its immediate ancestor, Deroceras Dudressleriy but on the other hand, Deroceras nodoyiijas is just as large. The septa, also, of the first named, which can hardly be said to be more complicated, cannot be claimed as simpler than those of Deroceras Dudresstcri. The varieties of Deroceras nmticum indicate that there is a tendency existing in the organization to flat¬ ten the sides, and render them convergent instead of divergent. Whether this is a governing one or not in the series cannot be C* O determined, and perhaps may never be, for it would not be at all surprising if these two species rej)resented all thai ever existed. It is well to remember, however, that this is the same tendency which finds a fuller expression in the series made by Dactylioreras communCy Holandreiy minulatuniy and Braufiianum, It is here, how¬ ever, only slightly shown, and the sides do not, in the most extreme variety, approximate, or become so convergent as to make the abdo¬ men acute, as in Dactyjdoceras Braunianum. The difference between the varieties of Deroceras muticum are, however, comparable, so far as the divergency or eonvergency of the sides are concerned, to the difference between the divergent sides of Cveloceras Desplaceiy and the convergent sides of DaciyUoceras Ilolandrei ov annulatum\ species undoubtedly distinct and easily de¬ finable, on account of numerous other characteristics. 4 Dereceras armatum. Amin. anhafi¿s Sow., ]Min. Conch., rol. i, p. 215, pl. 85. Dcroccni.^ armatum ITyatt, Bull. !Mus. Comp. Zoology, no. 5, p. 95. This pjn'cics has, from the earliest period, a very marked character. It is smooth much lonircr tlian the vouno: of Deroceras iJiuIreasieri. or than any of the Li])aroceratida', except ]\ricrota closely resemble ihose of Mtrroderoreraa Birchii^ — the large abdominal or sijdional cell in the centre of the abdominal lobe, with the adjoining superior lateral cells a2)]>arently stuck on to the side of the latter, — the height and extremely widcs])reading minor lobes, as well as the size of the suj)erior lateral lobes, and the dejuli. narrowness and mode of division of the superior lateral cells by the very 2)rominent minor 1o1h»s, — these decisive similarities, and the singularly close likeness of the very ornate outlines of all the lobes and cells show that there can be no doubt of the affinity of this sjn'cies for Mícrodcrocfra!^ Blrcliii \ but the single row of promi¬ nent spines indicates that, like Drroceras Dmiressieri, it is a generic derivative irom ^lirroderoreras Bircliii. And an examination of the young se])ta shows that at one stage they have the large cell in the centre (d* the superior lateral lobes so common in De roceras Dudres- sieri. Thus the si)ecies is ¡)ro2)erly and more directly a descendant of 1)< roceras Dudressieri with reversions to M'tcroderoceras Birchii. The. absence of a planicostan abdomen iit all stages places it in advance of Deroceras Dudressieri The geological j)Osition of the latter in the Lower Lias Obtususbett, and of this sjjccies on the bor¬ ders of the jMiddle Lias, confirm this view. Deroceras alternum. PenmocetKis altemium Hyatt, Bull., Op. cit., p. 85. The young of this sj)ecies is smooth, like the young of Deroceras armatum, the tubercles and pila* are developed slowly, as in that spe¬ cies, and the characteristics of the adults are similar, with ibid-like, 5 spined pilíP, and numerous erenulations between. Tlie abdominal cell is exceedingly larfre. Tlie abdominal lobe is very broad, and mueli deeper tlian the superior lateral lobes. The superior lateral cells are extremely narrow, and much cut into by the minor lobes. The superior lateral lobes have three wide spreading" branches and the intermediate cells therefore are vervlar're, and the inferior lateral cells vejy much attenuated at the top. These characteristics approxi¬ mate also to those of Deroceras muticnm. tliouiih not so elosclv as to tliose of Deroceras Davœl, while tlie shell is more like Deroceras armaium. It possesses just tlie same intermediate charaeter as Dero¬ ceras Dacœu hut is a very much smaller species, with the young more like Deroceras aimiatum, Deroceras Davoei. Anini. Darœi^ Sow., Min. Coneh., vol. iv, p. 71, ¡d. 350. Deroceras Daooei Hyatt, Rull. Mus. Corup. Zoology, no. v, p. 84. This species carries to excess the sparse distribution of tlie spines, the mere folds instead of pilte, and the multitudinous erenulations ob¬ served in Deroceras annatu/n. The spines unite these when they occur ill knots, but this is subject to the greatest variations, and the number of the spines varies considerably. They are not present in the young until a late period, and then occur pretty regularly, though much wider apart in some specimens than in others, and finally dis¬ appear suddenly as old age approaches. The advance of age is indicated not only in this way, hut by the irregularity of the erenu¬ lations. These, at intervals, are almost obliterated, anlain the sudden smootliness ; the animal aflfected by senility not being able to build and retain the erenulations, when tin* old rate of growth is suddenly resumed. The septa resemble closely those vf Deror/ras muticwu. The superior lalx^ral cells are, however, wider, the suj)erior lateral lobes deeply divided, the inferior lateral cells very narrow. The abdominal lobe is deeper than the superior laterals. Thus while more like Deroceras mnlicnm than Deroceras armafum in its septa and in the external eliaracteristies of the shidl, still the folds, instead of pilœ at each spinous node, the spines and the enmu- lations between tliem, are eliaracteristies uniting it with Deroctras % 6 armaturn. Tlie affinities tlicrefore are doubtful, though in my opinion it is nearer to the hitter than to the former, and is probably a spe¬ cies in the same series as Deroceras armatum. Deroceras muticum. P( ronoceras mulknm Hyatt, Bull. !Mus. Conjp. Zoology, no. V, p. 85. Arnm. 7/ni/¿cus D'Orl)., Terr. Jurass., Cej)h., pi. 80. This species has two well marked varieties, one closely resembling Deroceras Dudressieri in the breadth of the abdomen and the di¬ vergency of the sides; the other peculiar, with flat, or slightly con¬ vergent sides. The [ ila? and lower portion or base of the spines are not, as in Deroceras Dndresst^'ri, filled with shell, and they are verv distinct on the casts ; nor do the young or adults even, so far as I have seen, exhibit the j)lanic()stan abdomen. The young, in other respects, are very similar to the young of that species, but the pilae and the spines a[)i)ear at an earlier age. DT)rbignyV figure of the septa is excellent, and by comparison with those of Deroceras Duflressîerî many constant diflerences can be » observed, though the general characteristics are similar; the narrow, prominent, abdominal cell, the deep divisions, or minor three lobes at the bases of the superior lateral cells, the huge lateral cells on either side of the tops of the superior lateral lobes, and the extreme nar¬ rowness of the aperture at the top of the inferior lateral cells. According to OppeP this species is identical with part of Afnm. armafus c/ensi'nt/dus of Quenstedt, and is found in the Baricostatus- bed of the Lower Lias. This agrees admirably Avith its affinity for Derocf ras Dwíresskri^ which occurs in the next bed but one below the Obtusiisbed. Quenstedt (Der Jura, p. 105) also ¡daces this species with rarîcostatns in the Lower Lias. Tliere are, however, a large number of s¡)eeimens labelled Venary près de Semur, IMiddle Lias," by Boucault, ami these are associated with others which he has named snharrnatns*' but Avhich are merely the broad Deroceras Duflressieri'Viko A'arietv described above. % Deroceras nodogigas. Peronoceras tiodag'tgas Hyatt, Bull. Mus. Comp. Zoology, no. V, p. 85. Anm. nodogigas Quenstedt, der Jura, p. 125, pi. 15. Tlie Museum possesses one large, but indifferently preserved spec¬ imen of this species, which does nut enable me to throw much light ^ Die Jura formation. JaJirshefte fur Naturk. in Würt.» 1866, p. 209. 7 upon its affinities. Quenstedt asserts tliat it is distinct from Oppel's Amm. armatuii^ -wliicli appears to be plainty enough indicated by this specimen. The pilœ and spines have none of the irregular swollen and divided aspect of those of the typical armaius, but are closely similar to those of Derocera^ muticum. They are also mox'C numer¬ ous than in Deroceras armatiim. SERIES CYCLOCERATIN^. This series ^ exhibits some remarkably curious and interesting fea¬ tures. Like the Dactyloidae it does not increase the involution of the whorls, but merely the flatness of the sides, the narrowness and prominence of the abdomen. The approximation of the earliest existing species, Plafypleuroceras hrevispina to Microderoceras Birchiiy is very marked, aud very curiously the adult is almost identi¬ cal with the adult of Microceras latcecosta. From this species to the next but one, Cycloceras V(ddani is a considerable step. Not only has the character of the abdomen entirely changed from a rounded planicostan form to a keeled outline, but the young display this characteristic at an early age. The remaining species simply exaggerate these peculiarities. The geological succession does not strictly accord with the manner in which the organization of the species arranges them in the series. Plafypleuroceras hrevispina occurs in the lowest bed of the Middle Lias, Jainesonibed, and Cycloceras hipunctatus and Actceon in the next, or Ibexbed ; but Cycloceras Masseanum^ instead of ibllowing these species, as would be naturally expected, is associated with Plafypleuroceras hrevispina. This discrepancy will probably be ulti¬ mately explained, but at present it must be quoted as an exception to the usual law of agreement between zoological rank as determ¬ ined by development and geological succession in time. It is only fair, however, to state that the affinities of Cycloceras Masseanum were considered very doubtful before its geological position was as¬ certained. 1 This series was formerly described as a separate family in Bull. Mus. of Comp. Zoology, on accouut of the difiereiices of the adults. 8 PLATYPLEUROCERAS. Platypleuroceras brevispina. Amin, hrecf.ytina Sow., Min. Conch., yol. vi, p. 106, pl. 556. Ajnm. 7i(itr¡x (rotundus et oblongus) Quenstedt, Die Ceph., p. 85, pi. 4, figs. 16, 17. Plalyjih antceras laicecoMa Hyatt, Bull. Mus. Comp. Zoology, no. V, p. 92. Cyclorenis molare Hyatt, Bull. ]\Ius. Comp. Zoology, no, v, p. 92. In this species we find a more sinirular combination of the charac¬ teristics of Microderoreras and Micritceras, The young are for a » O lengthened period, until, in fact, they are oyer two inches in diame¬ ter, yery similar to Microderoceras BlrcIiH. Then the pilae begin to stretch across the abdomen, ibrniing a whorl closely identical exter¬ nally with that of the adult of M à •roce ras lafrrcosfa. The double line of tubercles is retained, howeyer, much more persistently, and is usually more prominent than in the latter. The septa are yery dis¬ tinct at the same age in this species, the sutures being much more complicated, with more prominent cells and deeper lobes, and also as in Mirroderoceras, yery dee¡)ly divided by three minor lobes. In the ailult this contrast is not so great ; the lobes and cells appear to spread considerably, and become more like those of the adult of 21 icroveras latœvosta. CYCLOCERAS. Cyeloceras natrix. Anna, Tiafrix Schlot. Petrefact. A mm. natrix Ziet. Verst. "Wurt., p. 5, pi. 4, fig. 5. CycJoreras mtfrix Hvatt, Bull. ]\lus. Comp. Zoology, no. V, p. 92. The adult ahme was observed, an/y'ssus Quenstedt, Die Ceph., p. 90, pi. 5, fig. 8. 9 Amm. Valdani Qnenstcdt, der Jura., p. 131, pi. 16, fij^s. 2, 3. Cydoceras Valdani Hyatt, Bull. ]\Ius. Comp. Zoology, no. v, p. 95. Amm. hipnnctnfuA Oppel, dor Jura., p. 280. The young of this species shows at an early age the keel and flat¬ tened whorl of the adult, thus showing no material ditference either in development or adult characteristics fi'om l^ropidoceras^ which genus is therefore suppressed. Cydoceras Actseon. Amm. Act<üon D'Orb., Terr. Jurass., Ceph., p. 232, pi. 61, figs. 1-3. Tropidoceras Acto'on Hyatt, Bull. Mus. Comp. Zoology, no. v, p. 93. This species is very closely allied to Cydocera!^ hl}>un('tatvm^ and I much doubt the j)ropriety of separating the two. The material at my command, however, does not permit me to examine the younger stages of Cydncern:^ T''ríWa/¿/, with the requisite fullness and accuracy. Cycloeeras -SSgseDii. Amm. JE!((C(i than anything else. The lateral pihe were biiiireated an«l coarsely tubereulated, as in that sj)eeies. It resembles 73os7;/ufv / in the in¬ crease of the single pila^ wliieli make their a])pearanee between tlu' bifurcated pilas. The latter do in ohl aire entirely "iive wav to the 1 O » O fc bifurcated, and in some adults are very numerous. It should be mentioned that Oppel iuund s})iues on the umbilical whorl of Sower- byV original, and the IMuseum specimen deseribed above lias one or two fine spines on what is ju-obably the eighth whorl. Dactylioceras Braunianum. Aniffi. Braudianus D'Orl)., Terr. Jurass., Ceph.. p. 527, ]>!. 1()4. JJaciyliocercix Brauíiidninn Hyatt, Bull. Mus. Comp. Zoology, tio. v, p. 95. The cast of this species is particularly interesting. At or just before the adult stage the cast does not receive any impression from the abdominal pilœ, and looks smootb, narrow and subangular. The pila? are jirominent laterally, and have very slight tubercles. 18 That the pil» are continued on the exterior of the ah(h>nien imti! a much later period than this, is %'ery j)r()})able. The youn^ of this species has. after the smooth stajre. a period with tuherculated pila\ during which the shell closely resembles the av/xa, a swelling. % "Appa, a knot. MUSEUM OF COMPARATIVE ZOOLOGY. 89 Hammatoceras insigne Hyatt. Amm. insignis Schlo^, Ziet., Yerst. Wiirt, p. 20, pL 15, fig. 2. Amm. insignis D'Orb., Terr. Jurass., Cepli., p. 347, pL 112. Amm. insignis Quens'dt, Die Ceph., p. 280, pi. 40, figs. 4, 5. Log. Ulirweiler and Gundershofen ; Coll. Dr. Krantz and L. de Köninck. Hammatoceras variabile Hyatt. Amm. variabilis D'Orb., Terr. Jurass., Ceph., p. 350, pi. 113. Log. Bantz ; Coll. Dr. Krantz. AMAIiTHEOIDJS. Pleuroceras* Hyatt. Abdomen flat, with keel and channels well defined; keel crenulated; channels vary from obsolete to deep and well defined. Pilae swelling be¬ low, tuberculated ; genicular bend prominent. Tubercles lateral, arranged along the line of envelopment. Umbilicus open. Ventral lobe narrow and but slightly deeper than lateral lobes; the latter unequally divided. Inferior lateral lobe small, shallow, equally divided. Superior lateral cell only partly exposed on the side, and to¬ gether with the inferior lateral unequally divided.! Sub-Genus No. 1. Sides of whorls exposed. Pleuroceras hawskerense Hyatt. Amm. Jiawskerensis Y. and B., Phil. Geol. York., p. 164, pi. 13, fig. 8. Loc. Yeovil ; Coll. H. W. Marder. Pleuroceras spinatum Hyatt. Amm. sjpinaius Brug., Ency. Meth., t. 1, p. 40, pi. 14. Amm. spinatus D'Orb., Terr. Jurass., L, p. 209, pi. 52. Loc. Whitby, Yeovil, Avallon, Quedlinburg, Coburg, Franconia, Banz, Gundershofen, and Canal du Danube ; Coll. Mus. of Stuttgart, Dr. Krantz, L. de Köninck, Bronn, Marder, and Boucault. Pleuroceras costatum Hyatt. Amm. costatus Schlot., Pet., p. 66, pl. 12. Naut. costatus Rein., Naut. et Argo., p. 87, figs. 68, 69. Amm. costatus Ziet., Verst. Würt., p. 5, pl. 4, fig. 7. Amm. costatus Bronn, Leth. Geog., pl. 22, fig. 12. * nXcvpóí', a rib. t Septa are described from one species only, — Pleuroceras spinatum. 12 90 BULLETIN OP THE Loc. England, Tours, Weimar, Bantz, Uhrweiler, and Bas-Rhin ; Coll, Sir C. Lyell, M. Boucault, Bronn, Dr. Krantz, and L. Agassiz. Sub-Genus No. 2. Sides of whorls partially covered and flatter, especially in the young. Pleuroceras pseudo-costatum Hyatt. Amm. costatus nudus Quens'dt, Die Ceph., p. 95. Amm. costatus nudus Quens'dt, Der Jura, p. 171, pL 21, fig. 3. Loc. Yeovil, Dumbleton near Cheltenham, Rogueport, Canal du Danube, Plateau de Larzac, Goslar, Gundershofen, Baiern ; Coll. Bronn, L. de Kö¬ ninck, and Dr. Krantz. Pleuroceras pseudo-spinatum Hyatt. Amm, costatus S]pinatus Quens'dt, Der Jura, p, 171, pi. 21, fig. 1-3. Amm. costatus spinatus Quens'dt, Die Ceph., p. 95, pi. 5, fig. 10. Loc. Vassy (Dep. Yonne), hiilhaud (Dep. de I'Aveyron), and Courcy; Coll. Boucault and L. de Köninck. Pleuroceras vittatum Hyatt. Amm. viltatus Phil. Geol. York., p. 164, pi. 13, fig. 1. Loc. Whitby ; Coll. Dr. Krantz. Amaltheus De Montfoet. Abdomen acute, keeled, and channelled ; whorls compressed laterally. Keel crenulated, well defined. Tubercles, when present, are in a single row along the line of envelopment. Umbilicus open, with the sides of the whorls exposed or only partially covered. Amaltheus gloriosus Hyatt. Amm. amaltheus coronatus Quens'dt, Der Jura, p. 169, pi. 20, figs. 9-12. Loc. Milhaud, Balingen, Pliensbach, Boll, and Oflerdingen ; Coll. L. de Köninck, Bronn, and Dr. Krantz. Amaltheus salebrosus Hyatt. Amm. amaltheus spinosus Quens'dt, Die Ceph., p. 95, pi. 5, fig. 4. Amm. amaltheus spinosus Quens'dt, Der Jura, p. 168, pi. 20, fig. 8. Loc. AVhitby, Semur, Strasburg, Miihlhausen (Bas-Rhin), Phensbach, Boll, Geyslingen, Balingen, and Gundershofen ; Coll. Dr. Krantz, Boucault, Bronn, L. Agassiz, and L. de Köninck. Amaltheus turgidus Hyatt. Amm. Amaltheus gibhosus Schlot., Pet. p. 10. Amm. Amaltheus gihbosus Ziet., Verst. Würt., p. 4, pl. 4, fig. 2. MUSEUM OF COMPARATIVE ZOOLOGY 91 Amm. paradoxus Stahl, Ziet., Verst* Würt., p. 15, pl. 11, fig. 6. Loe. Plateau de Larzac, Heitlingen, Boll, Lutzude bei Ilanover, Semur, Göppingen, Ofterfeld bei Goslar, and Pliensbaeh ; Coli. Mus. of Stuttgart, Dr. Krantz, L. de Köninck, Prof. Bronn, L. Agassiz, and Boucault. Amaltheus margaritatus De Hont, Amalikeus margaritatus De Montfort, Conch. Sys., p. 91. Amm, acutus Sow., Min. Coneh., v. 1, p. 51, pl. 1 7, fig. 1. Naut, rotula Kein., Naut. et Argo., p. 59, pl. 1, fig. 5. Amm. Stokesi Sow., Min. Conch., v. 2, p. 205, pl. 191, figs. 9, 10. Amm. clevelandlcus Phil. Geol. York., pl. 14, fig. 6. Amm. amaltheus Ziet., Verst. Würt, p. 4, pl. 4, fig. 1. Amm. margaritatus D*Orb., Terr. Jurass., L, p. 246, pl. 67. Loe. Whitby, Avallon, Semur, Milhaud (Dep. de PAveyron), Bas-Rhin, Lutzude bei Hanover, Eislingen, Rezingen, Wasseralfingen, Gundershofen, Mühlhausen, Boll, Ubstadt bei Bruchsal, Falkenhagen in Lippe, Balingen, and Göppingen ; Coli. Mus. of Stuttgart, L. de Köninck, Prof. Bronn, L. Agassiz, and Boucault. Amaltheus praestabüis Htatt. Amm. amaltheus nudus QuensMt, Der Jura, p. 167, pl. 20, fig. 4. Amm. amaltheus nudus Quens'dt, Die Ceph., p. 94. Loc. Robin Hood's Bay, Scarborough, Whitby, Mende in Lozère, Venarey près Semur, Milhaud, St. Cyr près de Lyon, Metzingen, Lutzude bei Hano¬ ver, Balingen, Geislingen, and Göppingen ; Coli. Mus. of Stuttgart, Dr. Krantz, L. de Köninck, M. Boucault, Prof. Bronn, and L. Agassiz. CYCLOCERATID-®. This family is remarkable for containing species which on the one side ally it with the Liparoceratidce^ and on the other with the higher HUdocera" tidcB, There is, however, a general agreement in the development and in the septal characteristics, which unite them in one family. The form is much more compressed laterally than in the Liparoceratidœ and the tuber- culations of the pilge separate them from the Ilildoceratidce. The young of Tropidoceras Actœon resemble the adults of Cycloceras Valdani^ and the young of the last in their turn are like the adults of Platypleuroceras latee- costa ; thus all three genera are closely connected by development. The abdominal lobe is of about the same depth as the superior lateral ; the latter is unequally divided into three minor lobes of variable length, and there is only one auxiliary lobe exposed to view on the side. Superior lateral cell is generally equally divided, and of great breadth. Inferior lateral, narrower and more prominent. 92 BULLETIN OF THE Platypleuroceras * Hyatt. Abdomen nearly as broad, or broader, than the dorsal side of the whorl. Pilae single, tuberculated, and extending across the rounded abdomen, as in Planiceras. The septa are minutely divided by minor lobes, very closely set. The abdominal lobe is deep ; sides abrupt. Superior lateral very narrow, deeper than the abdominal, and profusely branching. Inferior lateral not as deep as superior lateral, and of about the same breadth and general aspect. Abdominal cell large and serrated. Superior lateral very broad, about the same height as the inferior lateral. Platypleuroceras lataeeosta Hyatt. Amm. latœcosta Sow., Min. Conch., v. 6, p. 106, pi. 556. Amm. latœcosta Ziet., Verst, Würt., p. 36, pi. 27, fig. 3. Amm. natrijc-rotundus Quens'dt, Die Ceph., p. 85, pi. 4, fig. 17, Loc. Gegenberg, Hinterweiler, Welfiingen, Rentlingen, and Balingen ; Coll. Mus. of Stuttgart, L. Agassiz, Dr. Krantz, and L. de Köninck. Cyclocerast Hyatt. Abdomen rounded or keeled, not so broad as the dorsal side of the whorl. Pilae single, tuberculated, and not extending across the abdomen in the keeled species. Young smooth for the first two or three whorls, then be¬ come ribbed. Keel appears at an earlier stage of growth than the pilse. Septa not so minutely divided by minor lobes, and the large lobes less dendritic than in Platypleuroceras. The abdominal lobe of medium depth, and quite broad. Superior lateral of medium breadth and considerable depth. Inferior lateral about two thirds as broad and deep as superior lateral. One small auxiliary lobe exposed laterally. Superior lateral cell broad and depressed. Inferior lateral more prominent and narrower; small auxiliary cell exposed on the side. Cyclooeras molare Hyatt. Amm. natrix oblongas Quens'dt, Die Ceph., p. 85, pi. 4, fig. 16. Loc. Balingen ; Coll. L. de Köninck. Cyeloceras natrix Hyatt. Amm. natrix Schlot., Petrefaktenkunde. Amm. natrix Ziet., Verst. Würt., p. 5, pi. 4, fig. 5. Loc. Balingen and Rentliniren : Coll. L. de Köninck and Dr. Krantz. o Ö ' Cyeloceras Valdani Hyatt. Amm. Valdani D'Orb., Terr. Jurass., Ceph., p. 255, pi. 71. Amm. compressas Quens'dt, Die Ceph., p. 90» pk 5- fig- 3. * nXarvf, flat, and UXevpoi/j rib. Î Kv/cXoy, circle. MUSEUM OF COMPARATIVE ZOOLOGY 93 Amm. Valdani Quens'dt, Der Jura, p. 131, pi. 16, fig. 2-3. Log. St. Amand, Seinur, Balingen, Reutlingen, and Gagenberg ; Coll. Mus. of Stuttgart, L. de Köninck, M. Boucault, and L. Agassiz. Tropidoceras * Hyatt. 'Abdomen invariably keeled, much narrower than the dorsal side of the whorl. Pilae single, smooth or tuberculated in the same species, do not ex¬ tend across the abdomen in any species. Young are smooth for one or two whorls. Keel and pilae appear simul¬ taneously. Septa have a more complicated aspect than in the preceding genus, the minor lobes being deeper and more numerous. The abdominal very broad at the bottom, narrower above. Superior lateral lobe narrow, and about the same depth as the abdominal. Inferior nearly the same, but less branching than the superior lateral. One auxiliary lobe exposed on the side. Abdominal cell very broad. Superior lateral and inferior lateral cells very irregularly divided by minor lobes. One small auxiliary lobe exposed on the side. Tropidoceras Actseon Hyatt. Amm. Actœon D'Orb., Terr. Jurass., Ceph., p. 232, pi. 61, fig. 1-3. Loc. Semur and Schoppenstadt ; Coll. Dr. Krantz and L. de Köninck. Tropidoceras -Œïgseon Hyatt. Amm. Mgœon D'Orb., Terr. Jurass., Ceph., p. 234, pi. 61, fig. 4-6. Loc. Près de Semur ; Coll. M. Boucault. Tropidoceras Masseanum Hyatt. Amm. Masseanus D'Orb., Terr. Jurass., Ceph., p. 225, pi. 58. Amm. Masseanus Quens'dt, Die Ceph., p. 90, pi. 5, fig. 2. Loc. Scarborough in Yorkshire, Près de Semur, and Balingen ; Coll. Dr. Krantz, L. de Köninck, and M. Boucault. Upper Lias. DISCOCERATID-®. Ophioceras, Ophioceras Levesquei Hyatt. Amm. Levesquei D'Orb., Terr. Jurass., Ceph., L, p. 230, pi. 60. Amm. solaris Ziet., Verst. "VVürt., p. 19, pl. 14, fig. 7. Amm. radians quadratus Quens'dt, Die Ceph., p. 113. Loc. Niort, Salins, Heiningen, and Metzingen ; Coll. Mus. of Stuttgart, Dr. Krantz, and L. de Köninck. * XpoTTir, a keel. 94 BULLETIN OF THE DEBOCERATID^. Deroceras- Deroeeras minatum Hyatt, Loc. Plateau de Larzac ; Coll. Dr. Krantz. Abdomen depressed. Sides flattened or inclining toward umbilicus. Septal lobes and cells very simple, with but few minor lobes. Abdominal lobe broad and shallow. Superior lateral the same, and of nearly the same size. Inferior lateral pointed and very small. Superior lateral and inihrior lateral cells equally divided by minor lobes. Young are smooth for the first two or three whorls. Tubercles usually make their appearance on the tliird whorl and on the fourth ; these spread out into pilœ, and other untuberculated pilae arise between them. The pilas are often slightly de¬ pressed or concave along the siphonal line. Dereceras subarmatum Hyatt. Amm, suharinatus Sow., IMin- Conch., v. 4, p. 146, pi. 407. Amm. suharmatus Young and Bird, Geol. York., p. 250, pi. 13, fig. 3. Loc. Milhaud (De'p. de TAveyron) ; Coll. M. Boucault. Dereceras acanthepsis Hyatt. Amm. acanthopm D'Orb., Prod. Pal. Stratigraph., p. 247. Loc. Villebois (Dep. Ain) ; Coll. Prof. Bronn, DACTYLOID.®. Coeleceras. Cœleceras Greneuilleuxii Hyatt. Amm, Grenouillouxii D'Orb., Terr. Jurass., Ceph., pi. 96. Loc. Fontaine Étoupe and Fours in Calvados, Plateau de Larzac, Che¬ ville in Sarthe, and Semur ; Coll, Dr. Krantz, L. de Köninck, and M. Bou¬ cault. Cœleceras Desplace! Hyatt. Amm. Desplacei D'Orb., Terr. Jurass., Ceph., p. 334, pi. 107. Loc. Avallon (Dep. Yonne) ; Coll. M. Boucault. Cœleceras crassum Hyatt. Amm. crassus Phil., Geol. York., p. 12, fig. 15. Amm. crassus Quens'dt, Der Jura, p. 251, pi. 36, fig. 1. Amm. raquinianus D'Orb., Terr. Jurass., p. 332, pi. 106. Loc. Whitby, Milhaud, Laumiere, Cheville in Sarthe, St. Cyr bei Lyon, Plateau de Larzac, Villebois (Dép. de TAin), Salins (Dep. Jura), Semur, hlontpellier, St. Quentin, and Près de Verpillier ; Coll. L. de Köninck, Dr. Krantz, L. Agassiz, Prof. Bronn, and M. Boucault. MUSEUM OF COMPARATIVE ZOOLOGY 95 Cœloceras mucronatiun Hyatt. Amm. mucronatus D'Orb., Terr. Jurass., Ceph., p. 328, pi. 104, fig. 4-8. Loc. Whitby, Milhaud, Laumiere, Mende in Lozère, Donau-Main Canal, Salins in Jura, Près d'Avallon, Montpellier; Coll. Dr. Krantz, L. de Kö¬ ninck, Prof. Bronn, and M. Boucault. Dactylioceras * Hyatt. The abdomen is either equal in breadth, or less than the back, instead of being broader than, or equal in breadth to, the back, as in the preceding genera. The lateral pilœ in the adult are smooth and invariably single ; the abdominal pilae may be either bifurcated or single. The young have the same development as the young of Cœloceras crassumj but the tubercles are dispensed "with before the adult state is attained. (The tubercles are hardly distinguishable in the young of some species, such as Holandrei and Braunianum, especially on the fossil casts, but are, nevertheless, present in all the shells.) Septa do not differ materially from those of the preceding genus, except perhaps in the greater simplicity of the lobes and cells, which are hardly so close together or so complicated. Dactylioceras commune Hyatt. Amm. communis Sow., Min. Conch., v. 2, p. 9, pi. 107, fig. 23. NauU annularis Rein., Naut. et Arg., p. 79, pl. 6, figs. 56, 57. Amm, annularis Ziet., Verst. Würt., p. 14, pl. 10, fig. 10. Loc. Whitby, Boll, Arnberg, and Langenbrücken ; Coll. Dr. Krantz, L. de Köninck, and Prof. Bronn. Dactylioceras Holandrei Hyatt. Amm. Holandrei D'Orb., Terr. Jurass., Ceph., p. 330, pi. 105. _ ✓ Loc. Whitby, CheviUe in Sarthe, Fontaine Etoupe Fours in Calvados ; Coll. L. de Köninck and M. Boucault. Dactylioceras annulatum Hyatt. Amm. annulatus Sow., Min. Conch., v. 3, p. 41, pi. 222. Amm. annulatus D'Orb., Terr. Jurass., Ceph., p. 265, pi. 76, figs. 1, 2. Argo. anguinus Rein., Naut. et Arg., p. 89, No. 1, pi. 12, fig. 73, Amm. œquistriatus Ziet., Verst. Würt., pl. 12, fig. 5. Loc. Whitby, lUminster, St. Amand, Fontaine Etoupe Fours ; CoU. L. de Köninck and Dr. Krantz. Dactylioceras Braunianum Hyatt. Amm. Braunianus D'Orb., Terr. Jurass., Ceph., p. 327, pi. 104. Loc. Milhaud and Plateau de Larzac ; Coll. L. de Köninck and Dr. Krantz. * AaKTvXtor, a ring. 96 BULLETIN OF THE THYSANOIDJE. Thysanoceras Htatt. Thysanoceras fimbriatum Hyatt. Amm.ßnihriatuii Sow., Min. Conch., v. 2, p. 145, pi. 164. Loc. Pouilly in Côte d*Or and Plateau de Larzac ; Coli. L. de Köninck and Dr. Krantz. Thysanoceras Germaihii Hyatt, Amm, Germainii D'Orb., Terr. Jurass., Ceph., p. 320, pi. 101, Amm. interruptus Ziet., Verst. Würt., pl. 15, fig. 3. Amm. ohlique-cosiatua Ziet., Verst. Würt., pl. 15, fig. 4. Loc. Milhaud (Dep. de rAve}Ton), Semur, Pouilly in Côte d'Or, and Gundershofen ; Coll. Dr. Krantz, M. Boucault, and L. de Köninck. Thysanoceras articulatum Hyatt. Ajnm, articulatus Sow., De la Bêche, Geol. Manual, p. 276, fig. 63. Loc. Spezzia ; Coll. Prof. Brenn. Thysanoceras PhiUipsii Hyatt. Amin. PhiUipsii Sow., De la Bêche, Geol. Manual, p. 275, fig. 57. Loc. Spezzia; Coll. Prof Bronn. Thysanoceras cornucopia Hyatt. Amm. cornucopia Young and Bird, Geol. York., pi. 12, fig. 6. Amm. cornucopia D'Orb., Terr. Jurass., Ceph., p. 316, pi. 99. Loc. Semur, St. Quentin, près de Verpillier, Plateau de Larzac, and Milhaud (Dep. de TAveyron) ; Coll. L. Agassiz, M. Boucault, Dr. Krantz, and L. de Köninck. Thysanoceras torulosum Hyatt. Amm. torulosus Schub. Ziet., Verst. Würt, p. 19, pl. 14, fig. 1. Amm. scutatus Von Buch, Pet. remarq., pi. 8, fig. 1. Amm. torulosus D'Orb., Terr. Jurass., Ceph., p. 322, pi. 102. Loc. Plateau de Larzac, Zillhausen, Metzingen, Schömberg, and Dur¬ wangen ; CoU. Mus. of Stuttgart, Dr. Krantz, and L. de Köninck. Thysanoceras júrense Hyatt. Amm. Jurensis Ziet., Verst. Würt, pl. 68, fig. 1. Amm. jurensis D'Orb., Terr. Jurass., Ceph., p. 218, pi. 100. Amm. phyllocinctus Quens'dt, Der Jura. Loc. Semur, Plateau de Larzac, Milhaud (Dep. de TAveyron), Hechin- gen in Würtemberg, Uhrweiler in Elsass, Adnet bei Salzburg, Sondelfingen, Balingen, Reutlingen, Metzingen, and Gundershofen ; Coll. Mus. of Stutt¬ gart, M. Boucault, Dr. Krantz, Prof. Bronn, and L. de Köninck. MUSEUM OF COMPARATIVE ZOOLOGY. 97 Thysanoceras hireinum Hyatt. ÁTYim* hircinus Schlot., Pet, p. 72. Amm. hircinus Quens'dt, Der Jura, p. 280, pi. 40. Loo. Semur, Donau-Main Canal, and Mistlegau bei Bayreuth ; Coll. M. Boucault and Prof. Bronn. Hhacoceras L. Agassiz. Rhacoceras calypso Hyatt. Amm. calypso D*Orb., Terr. Jurass., L, p. 167, pi. 52, figs. 7-9. Loe. Plateau de Larzac, Monte de Aquasparta bei Cesi in Umbria, Mil- baud, Laumière, Digue in Basses Alpes, and Erba bei Como ; Coll. Dr. Krantz, L. de Köninck, and Prof. Bronn. The abdomen is broader than in R. heterophyllus, and the septa diifer- 4 ent, but, nevertheless, the varieties of this species which are devoid of the annular depressions caused by the permanent mouths, are frequently iden¬ tified with that species. RBaeoeeras heterophyUum L. Agassiz. Amm. heterophyllus Sow., Min. Conch., v. 3, p. 119, pi. 266. Amm. heteraphyllus D'Orb., Terr. Jurass., I., p. 339, pi. 109. Loc. Whitby, Boll, Vassy près d'Avallon, Erzingen (Dep. du Doubs), Balingen and Bruchsal; Coll. Dr. Krantz, M. Boucault, and Prof. Bronn. Rliacoeeras cylindricum Hyatt. Amm. cylindricus De la Bcche, Man. Geol., p. 275, fig. 55. Loc. Schoppenstadt ; Coll. Dr. Krantz. RKacoceras mimatense Hyatt. Amm. mimatensis D'Orb., Terr. Jurass., p. 344, pi. 110, figs. 4-6. Loc. Plateau de Larzac (Dep. de I'Aveyron) and Boll ; Coll. L. de Kö¬ ninck and Dr. Krantz. PHYMATOID.ŒÏ. Phymatoceras Hyatt. Phymatoeeras enervatum Hyatt. Loc. Plateau de Larzac and Villenotte près de Semur ; Coll. Dr. Krantz and M. Boucault. The abdomen is much flatter in this species than in the succeeding P. ro- hustum, and in the young the channels are deeper and more distinct. The increase of the radii of the spiral is also lees, and there are therefore a 13 98 BULLETIN OF THE greater number of whorls in specimens of the same size. The sides of the whorls are also less gibbous than in P, robustum^ and auxiliary cells differ¬ ently formed, being comparatively but slightly indented by the minor lobes, and the inferior lateral cell inclined toward the umbilicus, instead of being straight. o Phymatoceras robustum Htatt. Log. Plateau de Larzac, Milhaud, and Semur ; CoIL Dr. Krantz, L. de Köninck, and M. Boucault. Ammatoceras. Ammatoeeras insigne Hyatt. Amm. insignis Schub., Ziet, Verst. AVürt., p. 20, pl. 15, fig. 2, Amm. insignis D'Orb., Terr. Jurass., Ceph., p. 347, pi. 112. Amm. insignis Quens'dt, Die Ceph., p. 280, pi. 40, figs. 4, 5. Log. Gundershofen (Bas-Rhin) ; Coll. M. Boucault. Ammatoeeras variabile Hyatt. Amm. variabilis D'Orb., Terr. Jurass., Ceph., p. 350, pl. 113. Loc. Laumière, Salins, Plateau de Larzac, St. Julien de Croix in Saône et Loire, Besançon, Evrecy bei Caen, Boll, and Balingen ; Coll. Mus. of Stuttgart, L. de Köninck, Dr. Krantz, and M. Boucault. Pelecoceras* Hyatt. Having but one species of this genus, it would be exceedingly hazardous to give the generic characters. They will, however, probably be found to be distinguished by the peculiarly pointed aspect, shallowness and breadth of the lobes and cells ; the limits of the envelopment, which last is greater than in other genera of this family ; the acute form of the back, and the breadth of the whorls. Pelecoeeras attenuatum Hyatt. Loc. Plateau de Larzac, hlilhaud, and Besançon ; Coll. Dr. Krantz and L. de Köninck. Abdomen acute. Sides very broad and flat. Envelopment covers over one half the side of each internal whorl. Pilaî are curved forward on the abdomen. The young have no channels, and the development does not differ fi:om Amm. variabilis or Amm. insignis, except in the size of the young, the whorls of these not being proportionately so large or broad. All the lobes and cells are broad and shallow, especially the pointed ab¬ dominal and the serrated auxiliary cells. * HeXe/cvr, an axe. MUSEUM OF COMPARATIVE ZOOLOGY 99 HILDOCERATID-Ï3 * Hildooerast Hyatt. Abdomen keeled and channelled. Ribs large and broad. The young continue smooth throughout first whorl. Ribs, keel, and channels appear on the second whorl. The ribs are not preceded by a line of" tubercles, but begin as folds, bent much in the same way as in the adult, but with the ab¬ dominal bend inclined more toward the apex. The abdominal lobe is shallow and broad. Superior lateral much deeper than either the aodorni- nal or inferior lateral lobes, the last named very narrow and shallow, minor lobes small and pointed. Hildoceras bifrons Hyatt. Amm, bifrons Brug., Ency. Meth., Amm. No. 15. Amm. bifrons D'Orb., Terr. Jurass., Ceph., p. 219, pi. 56. _ X Loc. Whitby, Dumbleton, Dorsetshire, Fontaine Etoupe Fours, Poille in Sarthe, Laumière, Mende in Lozère, Verpillier, Milhaud, Plateau de Lar- zac, Cesi in Umbria, Mussy près de Semur, Chary près de Privas, Amayer sur Orne, Boll, and Metzingen ; Coll. L. de Köninck, Dr. Krantz, M. Bou- cault, and Prof. Bronn, Hildoceras Walcotii Hyatt. Amm. Walcotii Sow., Min. Conch., v. 2, p. 7, pi. 106. Amm. Hildensis Young and Bird, Geol. York., pi. 12, fig. 1. / Loc. Hlminster, Niort, Fontaine Etoupe Fours, Plateau de Larzac, Cesi in Umbria, Milhaud, Vieux Ponts, and Guadalaviar in Aragon ; Coll. B. M. Wright, Dr. Krantz, and L. de Köninck. Grammoceras í Hyatt. Abdomen keeled, but not channelled. Whorls flattened, laterally giving a discoidal aspect to the shells. Ribs finer and less prominent than those of Hildoceras. The young also continue smooth much longer, and chan¬ nels never appear ; they take, however, the same rounded form of the whorl. Septa differ but slightly from Hildoceras in the higher species, such as Grammoceras serpentinum ; and not all generically in the lower, such as Grammoceras striatidum. Grammoceras striatulum Hyatt. Amm. striatulus Sow., Min. Conch., v. 5, p. 23, pi. 421, fig. 1. Amm. Thouarsensis D*Orb, Terr. Jurass., Ceph., p. 222, pi. 57. Amm. radians depressus Quens'dt, Der Jura, p. 281, pi. 40. * Includes all the Falciferi proper with smooth pilae. t After St. HUda, J Tpapp7¡, a line 100 BULLETIN OF THE Loe. Whitby, ßobin Hood*s Bay, Milhaud, St. Julien du Cray in Saone et Loire, Niort, Plateau de Larzac, Près de Lyon, Boll, Keulwagen, Be¬ dangen, Heiningen, Aalen, Falkenhagen in Lippe, Metzingen, and Uhr¬ weiler ; Coli. Mus. of Stuttgart, L. de Köninck, Dr. Krantz, Prof. Bronn, and M. Boucault, Grammoceras radians Hyatt. A mm. radians Schlot., Pet. p. 78, No. 34. Naut. radians Bein., Naut. et Arg., p. 71, No. 17, figs. 39, 40. A mm. radians Ziet., Verst. Wiirt., p. 5, pi. 4, fig. 3. Amin. Uneatus Ziet., Verst. Würt., p. 12, pl. -Q, fig. 7. Amm. radians compressas Quens'dt, Die Ceph., p. 112, pi. 7, fig. 9. Loe. Niort, Plateau de Larzac and Camus in Cevenen, St. Cyr bei Lyon, Villebois in Ain, Salins in Jura, Milhaud, Mende, Besançon, Vaches Noires in Calvados, Uhrweiler, Falkenhagen, Boll ; Coll. Dr. Krantz, Prof. Bronn, L. de Köninck, and M. Boucault. Grammoceras aalense Htatt. Amm. aalensis Ziet., Verst. AVürt., p. 37, pl. 28, fig. 3. Amm. aalensis Quens*dt, Die Ceph., p. 114, pl. 7, fig. 7. Amm. aalensis D'Orb, Terr. Jurass., Ceph., p. 238, pl. 63. Loe. Trocester Hill, Milhaud, St. Vigor, St. Julien du Cray, La Verpil- lière in Ain, St. Quentin, Aalen, Ileiningen, Neuniarkt, Balingen, Mistle- gau, Arnberg, Wicsenthal, and Gundershofen ; Coli. Mus. of Stuttgart, L. de Köninck, Sir C. Lyell, L. Agassiz, Prof. Bronn, M. Boucault, and Dr. Krantz. Grammoceras costulatum Hyatt. Amm. costulatus Schlot., Pet., p. 78, No. 33. Amm. costula Krüg., Uhrwelt. Naturgesch., p. 27. Naut. Casilda Bein, Naut. et Argo., p. 68, pl. 3, fig. 33. Amm. radians costula Quens*dt, Die Ceph., p. 113, pl. 7, fig. 11. Loe. Aalen, Amberg, and Metzingen ; CoH. Mus. of Stuttgart, L. de Köninck and Prof. Bronn. Grammoceras serpentinum Hyatt. Amm. serpentinus Schlot., Pet., p. 64, No. 6. Argo serpentinus Bein., Naut. et Argo., p. 89, pl. 13, fig. 74. Amm. serpentinus Ziet., Verst. AVürt, p. 16, pl. 12, fig. 4. Amm. serpentinus D'Orb, Terr. Jurass., p. 215, pl. 55. Amm. Sirangewag.sii Sow., Min. Conch., v. 3, p. 99, pl. 25, fig. 1-3. Loe. Whitby, Somerset, Dorsetshire, Bannington, Milhaud, Fontaine MUSEUM OF COMPAKATIVE ZOOLOGY 101 Étoupe Fours, Thouars, Près de Semur, Vassy in Yonne, Amayer sur Orne, Boll, and Metzingen ; Coll. L. de Köninck, Dr. Krantz, M. Boucault, Prof. Bronn, JDuval, and Damon. Leioceras* Hyatt. Abdomen keeled, acute. Sides of the whorls flattened. Envelopment uniformly greater than in Grammoceras. The young differ, however, in being much flatter at the corresponding periods of growth. The lobes and cells, also, are less obtuse, shallower, and much more numerous. Leioceras lythense Hyatt. Amm. lythensis Young and Bird, Phil. Geol. York., p. 164, pi. 13, fig. 6. Loc. Whitby ; Coll. Prof. Bronn. Leioceras opalinum Hyatt. Naut. opalinus Rein., Naut. et Argo., p. 55, pi. 1, fig. 1. Naut comptus Rein., Naut. et Argo., p. 57, pi. 1, figs. 5, 6. Amm. primordialis Schlot., Pet., No. 7, p. 67. Amm. erratus Young and Bird, Phil. Geol. York., pi. 13, fig. 7. Amm. primordialis Ziet., Verst. Würt., p. 5, pl. 4, fig. 4. Amm. primordialis D*Orb., Terr. Jurass., Ceph., p. 235, pi. 62. Amm. opalinus Quens'dt, Die Ceph., p. 115, pi. 7, fig. 10. Loc. Robin Hood's Bay, Whitby, Trocester Hill, La Verpilliere in Ain, St. Quentin près Verpillièr, Szaflary, Amberg près de Goslar, Neuflen, Quedlinburg, Teufelsloch, Gundershofen, and Metzingen ; Coll. Mus. of Stuttgart, Dr. Krantz, Prof Bronn, L. Agassiz, and M. Boucault. Leioceras elegans Hyatt. Amm. elegans Sow., Min. Conch., v. 1, p. 213, pl. 94, fig. 1. Loc. Whitby ; Coll. Dr. Krantz. Leioceras complanatum Hyatt. Amm. complanatus Brug., Encycl., p. 38, No. 11- Amm. mulgravius Young and Bird, Phil. Geol. York., p. 251, pl. 13, fig. 8* Amm. elegans Phil. Geol. York., pl. 13, fig. 2. Amm. elegans Ziet., Verst. Würt., p. 22, pl. 16, fig. 5. Amm. complanatus D'Orb, Terr. Jurass., p. 353, pl. 114. Loc. Whitby, Lyme Regis, Villebois in Ain, Mussy près de Semur, Aval- Ion, Privas, Boll, and Ubstadt bei Bruchsal ; Coll. Prof Bronn, Dr. Krantz, M. Boucault, and Damon. * Aeîoy, smooth. 102 BULLETIN OF THE i Iieioceras discoides Amm. depressus Schlou, Pet., p. 80, No. 80. Amm. discoides Ziet, Verst. Wiirt., p. 21, pi. 16, fig. 6. Amin, depressus Ziet., Verst. AVürt., p. 7, pl. 5, fig. 15. Loe. MÍIhaud, Mende, Plateau de Larzac, and Balingen; Coll. L. de Köninck and Dr. Krantz. Bruguière (Encyclop., 1789), having described a different species by the name " depressus," Zieten^s name " discoides " is necessarily the correct name of this species. Leioeeras cumulatum Hyatt. * Amm. hicarinatus Ziet., Verst. "VVürt., p. 21, pl. 15, fig. 9. Loe. Milhaud, Laumière, Mende, Plateau de Larzac (Dep. de FAveyron), Montpellier, and Zillhausen ; Coll. L. Agassiz, L. de Köninck, Dr. Krantz, and M. Boucault. Zieten's " bicarinatus " differs specifically from Miinster's figure, Beit, zur Pet., v. 4, p. 138, pi, 15, fig. 30, and therefore it becomes necessary to adopt a new name for this species. Leioeeras coneavum Hyatt. Amm. concavus Sow., Min. Conch., v. 1, p. 215, pi. 94, fig. 2. Loc. Semur, Salins, and ITeiningen ; Coll. Dr. Krantz and M. Boucault. Leioeeras capellinum Hyatt. Amm. capelUnus Schlot., Pet., p. 65. Amm. capellinus Quens^dt, Die Ceph., p. 206, pi. 7, fig. 2. Amm. lythensis Uneatus Quens'dt, Die Ceph., p. 107, pi. 7, fig. 1. Loc. Metzingen and Holzmünden ; Coll. Dr. Krantz. From the Proceedings of the Boston Society of Natural History, Vol, XXII, April 4, 1883. GENERA OF FOSSIL CEPHALOPODS.i BY PROF. A. HYATT. Univalve shells maybe generally spoken of as cones, which may be either straight, curved, or coiled ; and the coiled may be either loosely coiled, or close coiled; either in the same plane, or a descending spiral. The shell-covered Cephalopoda are straight, ex. Orthoceras ; arcuate, ex. Cyrtoceras; loose coiled, ex, Gyro- ceras ; close coiled, ex. Nautilus. The larger number of the more ancient shell-covered Cephalo¬ poda are straight cones. These predominate in the Silurian over the arcuate, which are often merely varieties of species of the straight cones, as demonstrated by Barrande, and as may be observed in all good collections. The young of nautilian shells ai*e identical with the adults of the arcuate and gyroceran, and in different series repeat their forms, sutures, shell markings and 1 This paper is preliminary to a raonograph which will appear in the Memoirs of the Museum of Comparative Zoology. Hyatt.] 25i [April 4, the outlines of their whorl in transverse section. They are in succession, first, arcuate, then gyroceran, and lastly nautilian or close coiled. In several series genetic lines of adult forms may be followed, which lead by grad:»tion from arcuate, cyrtoceran forms to close coiled nautilian shells, the whole showing a connected series of transitions in the form and outline of section, sutures, structure, and position of siphon, and shell ornaments and apertures. In some cases these graded series are in accord with the chronologi¬ cal record, the straight appearing first, the arcuate either in com¬ pany with them or later in time, and the gyroceran and nautilian latest. "We cannot of course claim that such perfect evidence has been found even in the larger number of the following series. In some of them, certainly, it is not an over statement to say that the chronology of the evolution of form, the development of the individual, the gradations in the adults, and the general differen¬ tial characteristics all tell the same story, and are decisive for the opinion, that in all the larger series of shell-bearing Cephalopoda the nautilian shells belong to several distinct series and arose independently from straight cones through the intermedium of a graded series of arcuate and gyroceran or closely coiled forms. The generic terms, Cyrtoceras, Gyroceras and Nautilus are really only descriptive terms for the different stages in the development of an individual, and also the different stages in the development or evolution of the series of adult forms in time. In other words, each of these genera as now used, include representatives of all the different genetic series of Tetrabranchs, which are either young shells in the corresponding stage of growth, or adult shells in the corresponding stage of evolution. Our qualifications do not apply to the theoretical correllations, which may be observed between the development of the individ¬ ual in all its stages and the different forms of the group to which it belongb ; these are very nearly perfect. It is impossible to imagine the exact correspondence which exists between the transformations of an individual during its growth and the differ¬ ent adult forms of its own group, of the larva to the more ancient forms, of the adolescent stages to the lower forms of its own genus or family, except upon the theory of descent with modifi¬ cations. 1883.] 255 [Hyatt. The forms of the earliest fauna agree in their general aspect owing to the proximity of the septa, but they do not agree in structure, or in their embryos. The embryo in the Nautiloidea is a shrivelled protoconch, which may have been rounded at first but must have become shrunken and shrivelled after the animal passed out of it and into the apex ; it does not contain the siphonal coecum, and when broken away left a narrow cicatrix on the apex, the opening closed by a layer of shell. The umbilical perforation is an opening through the centre of the whorls of even the most completely coiled modern Nautilus, the hereditary mark of its uncoiled ancestry. The siphon may be near the venter, but the funnels rarely, if ever, break the continuity of the suture.^ The funnels of the siphon are simple posterior prolongations of the septa. The sutures are entire, they never have marginiil lobes and saddles or more than two lateral lobes ; the ventral lobe is usually undi¬ vided or simply V-shaped when it occurs,^ the dorsal lobe is, also, usually undivided, but may be divided in rare cases by saddles, the annular lobe when it occurs is undivided.® The siphon is variable in position, but the larger number of ancient genera have the siphon ventral or near the convex side It shifts in nearly all the series to near the centre, or dorsal side of the centre in the higher and often later occurring nautilian forms. The Aramonoidea have a globose protoconch, containing the coecum of the siphon, and when broken away it leaves the apex open. There is no umbilical perforation, except in the lowest and earliest of the Goniatitinae, the Nautilinidae, Some of this family have straight apices or young, others among them have arcuate and gyroceran stages, without the orthoceran, while the most closely coiled species in the adult are also close coiled in the larva and do not have arcuate and gyroceran stages. The higher Nautilinidae and all the succeeding genera have close coiled whorls in the earliest stages, with exceedingly rare exceptions, 1 Except, perhaps, in the Prochoanoids? 2 It may become divided by a broad saddle, the median saddle in rare cases. 8 Except in TrematoÍÍ¿SÉ and similar forms, where a median saddle is developed. Hyatt.] 256 [April 4, only occurring in tliC Devonian and Carboniferous, and the umbilical perfoiation is reduced to a mere depression on either side of the broad neck of the protoconch. The siphon is so near the venter that the funnel invariably breaks the continuity of the sutures with a slight lobe. The funnels of the siplion are not simple continuations of the septa, except in the Nautilinidae, and most forms of Goniatitinae. In some Goniatitinae, and all other Ammonoidea, a collar is foimed around the siphon in addition to the funnel, the collar being open and directed forwards. Tiie sutures are entire in most of the Goniatitinae, but in some species of the higher and later occurring carboniferous Goniatitinae marginal lobes and saddles begin to be formed, and in nearly all the remaining genera of later periods, these are char¬ acteristic. There are two lateral lobes w^hich arise in the Goniati¬ tinae by gradation from the simple lateral lobe of the Nautilinidae and in the higher forms an indefinite number of auxiliary lobes and saddles. The ventral lobe is undivided only in the Nautil- inidae!|^mT?n'aíf'^^ series of forms there is a median saddle, tlie sutures of this invariably divided by a minute funnel ^tL. The dorsal or inner side is occupied by a saddle in the^^^Taum- ^ inidae as is the case very generally among the lower forms of the different series of Nautiloids, all the remaining Goniatitinae have a broad dorsal lobe, wdiich is divided by small saddles as in the exceptional forms of the Nautiloidea, The annular lobe is absent among the Xautilinidae, as it is also universally in the lower forms of the different series of Nautiloîds.^ It is present, but undivided, in the remaining Goniatitinae, and is divided by a median saddle only in tlie liigher forms after the expiration of the Paleozoic. The appearance of a decided dorsal lobe correllates usually with the closer coiling of the whorls and the development of an impressed dorsal zone. This enables us to see, that the impressed zone is due to coiling, and also to define the gyroceran and nautilian forms. 1 I have found some rare exception.«?, adult stages of abnormal varieties with con¬ tinuous sutures over the venter as in the young of some Ammonites, Embryology of Ceph., Bull. Mus. Comp. ZooL, Vol. in, p. 110, fig. 1. These facts show that we are right in calling the minute central lobe on the venter, the funnel lobe. 2 Except some of the Endoceratidae according to Dewitz. 1883.] 257 [Hyatt. Thus, we can say, that a given form is still gyroceran though the whorls may touch in coiling, as long as the dorsum is rounded and gibbous, but if the dorsum has the impressed zone, it must be considered nautilian. This distinction enables us to add to the peculiarities of the Silurian fauna already noted by other authors (viz., the prevalence of straight, large siphoned forms, and those with septa closely approximate), the additional character¬ istic of the gi'eat raiity of true nautilian forms, M. Barrande has claimed that there was no approximation between the ancient forms of the Nautiloidea and Ammonoidea. We cannot understand the facts detailed above on any other supposition than the direct and independent derivation of the Nautilini from a straight cone. We think farther that this straight cone must have been a close ally and ancestor of the straight orthoceran-like Bactrites of the Silurian. This form agrees closely in all its characters with the young of the simplest known forms of Goniatites. The gyroceran and tubular whorls, and peculiar sutures and siphons of the young of Mimoceras are very similar to those of Bactrites. The seines of the Nautilinidae is, therefore, similar and parallel to that of any one series of the Nautiloids. It must have been independently derived from a straight cone similar to Bactrites. All the remaining ammonoids are more concentrated in development, and skip the orthoceran, cyrtoceran, and gyroceran stages of their evolution in time. They are evidently descendants of the close coiled ISTautilinidae and the evidence here is very strong that the whole order of Ammonoidea arose from a single organic centre of distribution, the Nautilini of the Silurian. The succession in time, the evidence of gradation in structure, and the development, exactly accord with this statement. Nautilinidae, Goniatites, triassic transition forms of Ammonitinae and the true Ammonites of the Jura form a perfect progressive series. The main difficulty in the way of the theory that Ammonoids and Nautiloids belonged to the same stock and were derived both from the same common ancestor laid in the assumed universal absence of a protoconch in the latter. We have found the proto- conch in several species of straight cones, and its absence in others can be readily accounted for. It was a useless hollow FROCEBDINOS B. 8. H, H. YOIh XXU. 17 DSCEMBBR, 1883. Hyatt.] 258 [April 4? appendage and probably also on account of its conchiolinouy struc¬ ture easily separated from the thicker calcareous shell of the apex. To close the evidence it only remains to point out the close aíBnities of the Bactrites for Orth, pleurotomum Barr. Syst. Sil. pl. 296 of Bactrites for the young of Mimoceras (Gon.) compressum Beyr. Sand. Verst. Nass. pi. 11, and the straight young of Agonatites (Gon.) fecundus, sp. Barr, Syst. Sil. pl. 11, fig. 4. The existence of the protoconch also removes a sei'ious objec¬ tion to the derivation of the Belemnoidea from the straight cones. We propose to remove another by homologizing the plug on the truncated cone of Orthoceras "with the guard of tlie Belemnitcs. We find that the central trace compares "with the pseudosiphon of tlie plug, and that the bilateral formation of the plug is similar to that of the guard. This indicates to our mind, not tlie exist- ^ y ence of two secreting organs like the arms of Argonauta, which O O o ' stretclied back over the shell of Orthoceras, as supposed by Bar¬ rando, but on the contrary an organ probably the homologue of the dorsal fold of the mantle in Nautilus. This could readily have been larger than in Nautilus and covered in the whole shell, and been divided into two secreting lobes at the posterior end. We are thus able to account for the inclosure of the shell among the Belemnoidea and the deposition of the guard, for the openness of this sac as shown by Branco in the transitional form, Aulacoceras of the Trias, and for its final closure as permanent sac among typical Belemnitcs without calling to our aid any ex¬ traordinary modifications of the known organs of Nautilus. The succession here would be Orthoceras, Silurian, Aulacoceras and Belemnitcs, Triassic, The Sepioidea appear to be connected Avith Orthoceratites through Gonioceras, Avhich resembles the broad internal shell of the Sepia ofiicinalis in the striae of growth and differs from all other forms of Nautiloidea in this respect. It has also septa whose outlines approximate to the outlines of the calcareous layer in the interior of sepia shell, or cuttle bone. Gon. occi- dentalis Hall, Rep. Geo. Surv. Wiscon., 1861, p. 47, has shell and septa, and the outlines of the form are also similar to Sepia, being broadly fusiform, and much compressed. The loss of the proto¬ conch can be accounted for in these forms in the same way that we can account for the resorption and loss of the siphon and 1863.] 259 [Hyatt degradation of the septa to a mere succession of calcareous layers- Namely, the shell having become internal and these parts being useless they gradually disappeared. They were first degraded, and then lost out of the roll of hereditary char¬ acters, the shell itself following in the same train, and disappear¬ ing at last in the higher and more specialized Octopods. We thus have Gonioceras in the Silurian, Paleoteuthis a true Sepioid, Devonian ; Loligidae or Teutliidae, Jurassic ; Octopods recent. To clinch this evidence wo can refer to tlie work of Lankester, " Development of the Pond Snail," in which he shows that the pen sac is not an enlai-gement of the shell gland, but must have been derived from a secondary sac formed by some extension of the mantle, which inclosed the shell and became a permanent addition to the organization, and we differ from this author only in imairininG: this inclosure as due to the extension of the dorsal O O flap of the mantle, as in Nautilus, instead of to two flaps of the mantle as in Aplysia, We reirard these two orders as distinct from each other but as O forminsr a division toirether, which we can designate as a sub-class O O ' O under the name of Dibranchlata equivalent to Professor Owens order of the same name. It is, however, not yet clear that they ai'ose from the same type among the straight cones, though that they both came from some straight Orthoceras seems to be indicated by all the evidence now in our possession. If now we return to the Nautiloidea and Ammonoidea we find these two orders to be unitable as having external shells a com¬ mon form of embryo and chambered shell, the chambers pierced by a siphon; that they possess similar structures, similar imbri¬ cated layers in the shell, and similar external deposits ; that they exhibit parallel series of forms in the independent reproduc¬ tion of the nautilian shells out of straight cones, and of the parallel modifications described above among the Goniatitinae. There is, therefore, every probability that they have been pro¬ perly united by Prof. Richard Owen as Tetrabranchs. We, however, prefer to consider this a sub-class with two orders according to tlie classification proposed by Prof. Louis Agassiz. We have, after much observation, found that genetic affinities on a large scale are best exhibited by the siphon, particularly by the funnels of the septa, which are more invariable than any other parts of the shell. Hyatt.] 260 [April 4, Following out the history of these parts the Nautiloids can be subdivided into the following general genetic groups, [1] Holo- choanoida, those having long funnels which completely close the intervals between the septa. We can subdivide this group into Prochoanites, those with funnels of septa turned forwards, ex. Bathmoceras, Metachoanites, those with funnels turned backwards and completely closing up the walls of the siphon, which has no intervening connective wall, ex. Endoceras, and Trocholites. [2J Ellipochoanoida, those with short funnels and the siphon completed by an intervening connective wall of distinct structure from the septal funnels. There are many of the Orthoceratites which have funnels of considerable lon the Goniatitinae except the Nautilinidae^^nd aÖ the Ammoni- tinae, and the funnel lobe is elevated upon it. An important factor in this classification is the dorsal suture, and we find that the position of a genus may often be determined in any given series by the peculiarites of this part ; whether it is present or absent, and whether it has, or has not a small annular lobe, or " spindle lobe," or a small saddle in the median line of the dorsum. All the series, with few exceptions, begin in time with arcuate forms which have dorsal saddles, and are succeeded by nautilian shells with dorsal lobes, and then these acquire the median annular lobes ; if they retain saddles on the dorsum, the dorsal lobe j invariably appears in descendants, but is apt to be divided by a small saddle in place of an annular lobe. So far as we know, the annular lobe appears in no species earlier than the Devonian.^ The " endosiphon," here spoken of for the first time by that name, is the internal tube long known in Actin- oceras, and lately demonstrated in Piloceras by Dawson, as hav¬ ing its own proper walls. To this we can add a similar apparatus observed in two good specimens of Endoceras, and also noted by the author in some specimens of Sannionites.. Among Nautiloidea there are no series traceable directly to arcuate forms after the expiration of the Carboniferous. This is the common story, and we can see that the series must have arisen very rapidly during the Paleozoic, branching out on every side from the common ascending trunk of the straight and arcuate forms. The same is true of the Ammonoidea in the Silurian, but only one short series, the Nautilinidae, arises from the common trunk of the straight cones. The close coiled shells of this series O become the stock form for the whole of the Ammonoidea. The Nautiloidea of the Mesozoic are all nautilian forms and their genetic series do not present the rapid changes of form observed in the Paleozoic, they are all close coiled and have as observed by M. Barrande small umbilical perforations. This same statement applies also to the Ammonoidea, when near their point of origin in the Silurian their forms are very quickly evolved, but are much less quickly evolved after this period. The smaller genetic groups in the Paleozoic are distinguished 1 These statements apply only to Nautiloids. See description of Goniatitinae Nautilinidae, and Agoniatites. Hyatt.] 262 [April 4. by differences between the sutures which are marked and decided by structural distinctions. Thus the groups of Clymeninnae and Goniatitinae, differ widely in their sutures and position of siphon and smaller groups have also decided structural differences. In later times the families and in fact the whole of the Ammo- nitinae are very similar in their sutures. There are, however, many genetic series, in the Jura families, which can be distin¬ guished by the minor details of the outlines of the sutures, but these distinctions are not so marked as in the Paleozoic, and the form of the whorl in section, and costations and ornaments of the shell are decidedly characteristic. In other words the field of variation is structually decidedly narrower, in the Mesozoic than in Paleozoic, whether we con¬ sider the Nautiloidea or Aramonoidea. We have observed the same phenomena repeated in each forma¬ tion and in the mode of appearance of all the genera and families. These groups originate suddenly and spread out with great rapidity and in some cases as in the Arietidae of the Lower Lias are traceable to an origin in one well defined species which occurs in close proximity to the whole group in the lowest bed of the same formation. These facts and the acknowledged sudden appearance of the larger number of all the distinct types of invertebrata in the Paleozoic, and of the greater number of all existing and fossil types befoi*e the expiration of Paleozoic time, speak strongly for the quicker evolution of forms in the Paleozoic and indicate a general law of evolution. This we think can be formulated as follows, types are evolved more quickly and exhibit greater structural differences between genetic groups of the same stock while still near the point of origin, than they do subse¬ quently. The variations or differences may take place quickly in the fundamental structural characteristics, and even the embryos may become different when in the earliest period, but subsequently only more superficial stmctures become subject to great variations. During this investigation we have been able to add to the facts we have already brought forward in support of the law of acceleration, or as we now prefer to designate it, the law of con¬ centration of development. All more generalized or lower types have a direct mode of development and the more specialized or complicated progressive types have, when at the acme of their 1883.] 263 [Hyatt. development, a more indirect mode of development. The types which are descended from these last have often a mode of develop¬ ment which in many forms is an apparent return to the direct mode of development again. The first two modes occur in the pi'ogressive series, the last can occur only in the highly retrogressive or degraded forms and con¬ sists of the following stages, to which naturallists acquainted with the life histories of modern parasites will easily find parallels. The degraded uncoiled forms of the Nautiloidea and Ammonoi- dea, wherever they occur, whether in the Silurian or in the Creta¬ ceous, invariably have close coiled young, showing that they were the offspring of close coiled or nautilian shells, that is of progres¬ sive forms which have themselves been evolved from a series of straight, arcuate, and gyroceran predecessors. Their uncoiling then is a truly retrogressive character, and this tendency to retro¬ gression is inherited in successive forms in several series. Their whole structure is finally affected, the whorl is reduced in size, and the complication of the sutures and shell at all stages of growth is degraded, until in their development only the close coiled young remain to testify to their exalted ancestry. In other words the forms inherit the degraded characteristics at such an early stage that it effects their whole life except the earliest stages. If Ave examine any of the pi'ogressive series we find that characteristic modifications or variations tend to appear first in the adults, then in successi\^e forms they appear at earlier stages, and finally disappear altogether or become embryonic, and this is the case also with the degraded characteristics, and doubtless when carried far enough even the last fortress of the ancestral characteristics, the larval stages would be invaded and the shell become completely uncoiled and perfectly straight and cylindrical from the earliest age. We have found specimens of Crioceras, in which only a part of the first whorl was close coiled and the embryo of the Baculite, the straight cone of the Cretaceous, and Jurassic Ammonoidea still remains unknown. We have, there¬ fore, in the life of a series heredity acting in such a manner that new characteristics are being continually introduced into the adult and adolescent stages to replace the ancestral ones Avhich have dis¬ appeared or been crowded back into the earlier or larval stages. Hyatt.] 264 [AprU 4, It is an undoubted fact, as shown by the miter and especially by Barrande and Dr. Branco, that the embryo itself has varied comparatively little throughout time in the Ammonoidea, Nautil oidea, Beleranoidea, and Sepioidea. But these statements do not apply to the earliest stages in the evolution of these types. During these earlier stages, when they all branched out from the common stock, the embryos of the Ammonoidea and Nautiloidea became quite different from each other, the embryos of the Belemnoids remained like those of the Ammonoids almost exactly similar to those of the Nautilini as shown by Chaimas and Branco, and finally in the Sepioidea the pro- toconch or embryonic shells changed more completely and soon disappeared. Attention is particularly called to this remarkable fact in the history of the evolution of these forms, that the separa¬ tion of the orders took place rapidly, and in the embryos as well as in the adults near the origin of the orders, and the comparative invariability of the embryo was confined to the subsequent history of these types after separation. We have here no space to discuss the apparent reasons for these changes, but we have been able to explain the mode in which they take place. The mode in each case is the earlier or concentrated development of ancestral characters, which as we have said follow the same paths, whether progressive and tending to preserve the characters of the type, or retrogressive and tending to destroy the characters of the type. We mention the law of concentration of development because in looking at the young in the usual haphazard way, naturalists often do not find the strong marks of afiinity which the ordinary modes of studying lead them to anticipate. The law of concen¬ tration leads to the disappearance of important characteristics often even in short and comparatively small series. It acts fre¬ quently within a small group like the Arietidae, so that the later larval and adolescent stages are exceedingly unlike the same stages in very nearly related species in the same family. Unless they are willing to take a small well characterized group and follow out all its transformations they cannot hope even to understand the remarkable phenomena which are shown more or less in the history of every complete series. Slaves of the embryological lamp consider that they must asso- 1883.] 265 [Hyatt. ciate all forms which have similar embryos, and dissociate in classification all forms having different embryos. As a matter of experience, the surest guides of affinity are the adult gradations of forms. These show that the Nautiloidea and Ammonoidea with comparatively distinct embryos are nevertheless closer related than the Belemnoidea and Ammonoidea which have precisely similar embryos, and Sepioidea and Belemnoidea which have very distinct embryos must also closely be affiliated. The embryos of all these must have been precisely similar at their origin, but they afterwards became varied in the different orders, and we cannot lay down any hard and fast rule by which the embryo becomes an invariable criterion of affinity. We think there is ample reason in the structures of these shells themselves for the embryonic differences, and that it is possible to reconcile them with the affinities indicated by the gradations observed between the adults. These reasons which we have space only to allude to here consists in a series of correlations which are plainly apparent between the adult structures, and the habits of the animals, and the tendencies which the habits have to change the adult structures, and then by the action of the law of concentra¬ tion in development to change even the embryos, either quickly in time when the habits are widely changed, or more slowly when they vary but slightly with the progress of time. The evolution is a purely mechanical problem in which the action of the habitat IS the working agent of all the major changes ; first acting upon the adult stages as a rule, and then through heredity upon the earlier stages in successive generations. Thus in the open fields of the periods of their origin they expanded into their different habitats, varying to accomplish this purpose with great rapidity, but once in their appropriate habitat inducements to change or open fields became rarer, and we get as a result comparative invariability. As time rolled on and the earth became more crowded, the variability was reduced to less and less important structural changes, except in the retrogressive types. These exceptions are our best proofs of the action of the habitat. The chanv, a column. Hyatt.] 276 [April 4, same author. It includes all those longicone species in which the ridges become spiny, or are roughened by the prominence of tlie transverse striae or ridges. Silurian ? Devonian, Carboniferous. Spyroceras,^ nobis, includes the longitudinally ridged longi- cones, wdiich at some stage of their growth are also annulated. The annular costae are usually large rendering the outline sinuous. The longitudinal ridges are present in the young, and the annular costae are developed later. Includes groups 5, 6 of M. Barrande. Type, Spy. (Orth.) crotalum,sp. Hall. Mus. Geol. Survey, Albany, N. Y. Silurian and Devonian. Dawsonoceras,^ includes forms like D. (Orth.) pseudo calami- teum, sp. Barr., pl. 286, and others which have longitudinal ridges in the larva and are annulated, but devoid of ridges in the adolescent and adults. The type is related to the series with large annulations and frilled transverse striae, sometimes with longitudinal ridges, though the young in D. (Orth.) dulce, sp. Barr., pl. 275 have no longitudinal ridges. The apertures have flaring lips as in Halloceras. Type, Daw. (Orth.) annulatum, Mus. McGill College, Montreal. Rizosceras,^ nobis, includes the straight cones figured by M. Barrande, Vol. 2, pis. 185-195, having simjde sutures, and septa, and banded shells, whorl in section elliptical. The form is remarkably short, and increases very rapidly, the living chamber short, and widely flaring, with shallow ventral sinus. The siphon is variable in position, but is rarely near the centre. Type, Riz. (Orth.) indocile, jil. 185, figs. 1-6. We include in this grouj^. also, such forms as are intimately connected with Rizosceras, like Riz. (Cyrt.) corniculum, pi. 121, sp. Barr., and (Cyrt.) apertum sp. Barr. pl. 146. The larger part of the forms figured by Barrande, pis. 1-158, belong to our genus Maelonoceras, and to various other groups, pis. 149-153, however, exhibit almost exclusively cyrtoceran, forms with rizosceran affinities. An extreme form of this genus is the Riz. (Phiagm.) imbricatum, sp. Barr. pl. 175. This group can also be subdivided according to the character of the siphon, 1 SmpCs, a basket. 2 Dedicated to Dr. J. W. Dawson of Montreal, ® PCso, a root. 1883.] 277 [Hyatt whether nurariiuloidal, or tubular. Its direct connection with Geisonoceras is evident, as well as the intermediate nature of all its characteristics with reference to the Gomphoceratklae. In this genus, as in Sactoceras, etc., the nummuloidal siphon is preceded by a tubular siphon in the earlier stages of growth, and this shows that we are justified in deriving this group from tubular siphoned orthoceratites, and considering it as transitional to the nummuloidal siphoned group of the Gomphoceratidae. Sometimes the nummuloidal siphon appears to juecede the tubu¬ lar, but in the cases figured by Barrande, and in others, which we have studied, the nummuloidal character was an adult peculiarity, and the so called tubular character was the result of degradation or shrinkage, due to old age. Gomphoceratidae, [Equilobates.] The shells have apertures with median saddles in the dorsal outlines, and, therefore, an equal number of lateral lobes. Acleistoceras,^ nobis, includes brevicone forms with a fusi¬ form shape, and partially contracted living chamber. The aper¬ ture has large ventral sinus, and a dorsal saddle, and is only slightly smaller in diameter than the living chamber, and the out¬ line is usually subtriangular. The siphon remains ventral, and the form in section is an oval with the dorsum broader than the venter. Silurian, Devonian, and Carboniferous. Type, Acl. (Api.) olla, Saern., Paleontogr. Vol. 3, pi. 19. Mus. Comp. Zool., Cam¬ bridge. Apiocera has been used for insects by Westwood, and the original figure of Apioceras by F. de Waldh. Bull. Soc. Nat. de Moscow, Vol. 17, p. 779, pi. 19, fig. 1, is not identifiable. Gomphoceras, Sow., Murch. Sil. Syst. Vol. 3, p. 620, includes all the straight and arcuate forms, which have symmet¬ rical T shaped apertures ; and, therefore, includes Phragmoceras, and the groups Dimorion and Dimeres of M. Barrande. Tetrameroceras, nobis, includes Silurian species having four lateral sinuses, and equals to the groups Tetramorion, and Tetrámeros of Barrande. Type, Tet. (Phrag.) bicinctum, sp. Barr. pl. 51, * AkXcuttos, open. Hyatt.] 278 [April 4, Hexameroceras, nobis, includes Silurian species having six lateral sinuses in their apertures. Type, Hex. (Pbrag.) Panderi, sp. Barr. pl. 48. [Inequilobates.] This series differs from the above in having an azygos dorsal sinus in jtlace of the median dorsal saddle of the preceding genera. All the genera are, so far as "vve know, Silurian. Trimeroceras, nobis, has only two lateral sinuses in addition to the median sinus. Type, Tri. (Gomph.) staurostoma, sp. Barr. pl. 73. Fentameroceras, nobis, has four lateral sinuses. Type, Pen. (Gomph.) mirum, sp. Barr. )»1. 82. Septameroceras, nobis, has six lateral sinuses. Trimeroceras was included by M. Barrande in his group of Trimorion, and Pentameroceras in Pentamorion. Septameroceras is founded upon a species in the Mus. Geol. Surv. Can., S. (Gomph.) inflatum, sp. Bill. Mesoceratidae. In this group we include all those brevicones whose short, con¬ tracted, bulbous, living chambers, and singular habit of truncat¬ ing their shells, and general tendency to flatten the abdomino- dorsal diameters of the apertures, and imperfect septa in the liv¬ ing chambers render them very distinct as a group from all other forms except the Asoceratidae. Mesoceras, Barr. Syst. Sil. Vol. 2, Text 6, p. 198, includes but one species. This has a much flattened aperture and very slight ventral sinus. It is an Acleistoceras without the vertical arm in the T shaped aperture. Billingsites,^ nobis, includes Silurian species having stout cones, almost globular on account of their truncation and which have dumb-bell shaped apertures, without ventral sinuses. Type, B. (Ascoe.) Canadense, sp. Bill. Kep. Prog. Geol. Can. 1853-56, p. 310, Mus. Geol. Surv. Can. This species shows that M. Bar¬ rande is in error, in supposing that the large posterior part of the living chamber can be considered as the siphon. The three last 1 Dedicated to the memory ofE. Billings. 2 A fact already noted by Blake, British Foss. Cephalop. 1883.] 279 [Hyatt septa are directly continuous with the septa on the dorsal side or the living chamber,^ and these are merely large dorsal saddles. Associated species have tliese dorsal saddles separated as in M. Barrande's typical Ascocei'as. The direct derivation of Mesos- ceras from Acleistoceras can hardly be doubted after comparing the apertures and the forms of the short living chambers. Ascoceratidae. European Silurian foians generally have annulated whorl, long living chamber, constricted near the aperture. The apertures are either open or obscurely Y shaped, the two arms of the Y being divided by a dorsal saddles, and the basal arm is the ven¬ tral sinus. Aphragmites, Barr. Syst., Sil. Vol. 2, Text 1, p. 366, is regarded by that author as a form of Ascoceras, which has re¬ sorbed the imperfect septa in the living chamber. We, however, much prefer this eminent author's first opinion, that it is a dis¬ tinct genus with simple septa and sutures. Ascoceras, Barr. Syst. Sil. Vol. 2, Text. 1, p. 334, includes certainly two groups, one with annulated shells, like those of Aphragmites, and bearing relations to this genus similar to those which Billingoceras has to Mesoceras; and one with smoother or banded and striated shells, which have apertures similar to the typical Ascoceras and are in the same genus. Griossoceras, Barr, ibid., p. 372, has a very slender whorl, and the obscurely Y shaped apertures described above. The species are Silurian, and the forms and markings of these fossils seem to indicate clearly derivation from an annulated stock like the Cycl. (Cyrt.) residuum sp. Barr. })L 286, which has similar attenuated, annulated whorls, but open apertures. Ophid loceras, Barr., includes Silurian shells closely coiled in the larval and adolescent stages and open in the later stages. The costated, compressed whorls have some resemblence to those of Ascoceras and the aperture is closely similar to Glossoceras. The shells are truly nautilian in the young and are evidently pa¬ thological derivatives of some ancestral nautilian form. The sutures ai*e straight and the abdomen has a blunt keel. They ap¬ pear to be the survivors of ancient ornamented series of costated shells. We place thern provisionally near Ascoceras on account of the Y shaped apertures and form of whorl and costations. Hyatt.] 280 [April 4, Maelonoceratidae. This family includes shells with whorls in section ovate, very short living chambers, often more or less compressed or with con¬ tracted apertures. The compressed apei'tures tend to become fusiform, and the contracted apertures become pear shaped. They are of smaller size than the Gomphoceratidae, and do not grade into that group, but have their own radical, open-apertured, cyrtoceran forms, which are included in the genus Maelonoceras. The sutures have ventral, and dorsal saddles, and lateral lobes. The siphons are near the venter and nummuloidal. Maelonoceras,^ nobis, includes Silurian species with arcuate cones, whorl in section compressed, ovate, the dorsum wider than the venter. The siphon is near the venter. The sutures have ventral and dorsal saddles, and slight lateral lobes. The living chambers are short, and the apertures vary from entirely open and partially subtriangular to contracted and pear shaped. Type, Mael. (Phrag.) praematurum sp. Bill. Can. Nat. Vol. 5, p. 173, fig. 19. A close ally of this is Mael. (Cyrt.) discoideum, sp. Barr. Syst. Sil. pl. 135. Mael. (Cyrt.) Metellus, sp. Bill. Pal. Foss. p. 191, fig. 175,176, is an open-apertured species. Mus. Geol. Surv. Can. The genus may have arisen from the same common stock, in Rizosceras but is cei'tainly not a direct derivative of Acleistoceras. Oonoceras,^ nobis, includes seides of European forms, which seem to arise from arcuate forms with open apertures, but more elongated cones. They are annulated, and have even shorter and more compressed living chambers in proportion to their longer shells. They may either retain the open aperture, or produce a fusiform' outline in the opening. Oon. (Cyi*t.) acinacies, sp, Barr, pl. 118, Giebeli, pl. 123, exile and letheum, pi. 124, are examples of cyrtoceran, and probably gyroceran foi*ms, and Oon. (Troch.) priscum and clava, pl. 12, oxynotum, pL 14, anguis, pi. 16, are examples of the closer coiled species There exist, doubtless in other localities, congeneric, close coiled, symmetrical shells. These Silurian forms lead into those with fusiform apertures, such as Oon. (Cyrt.) multiseptatum, Roem. Paleontogr. Vol. 3, pi. 6, fig. 2, and Oon. (Phrag.) sub-ventricosum, sp. D'Arch. et Vern. Geol. Trans. Vol. 6, pi. 30, of the Devonian. 1 M-qXov, a goat. 2 ftov, an egg. 1883.] 281 [Hyatt. StrGptoceras, Bill. Geol. Surv. Can. 1866, Antic. Foss. p. 88, fig, 28, appears to be identical with Acleistoceras, but the aper¬ tures are more like those of Maelonoceras. The form, however, is very distinct from both of these genera, and the living chambers one-third of the length of the whorl. This would not be impor¬ tant in most series, but in this one it is an extraordinary variation, and is perhaps an indication of essential differences. Mus. Geol. Surv. Can. Cranoceras,^ nobis, includes arcuate Silurian species, which have depressed elliptical whorl in section, and are very closely allied to Maelonoceras. T1 e sutures straight, or with dorsal and ventral saddles, and lateral lobes. The siphon is near venter. In the later stages of more curved forms slight dorsal lobes are developed. Cran. (Cyrt.) hospitale, sp. Barr. pl. 151, nigrum, ibid., pl. 127, Turnus, ibid., pl. 483, 484, of the Silurian connect intimately with the Devonian type (Cyrt.), depressum, sp. Goldf,, D'Arch. et Vern. Trans. Geol., Soc. Vol. 6, pL 29, fig. 1, Schultze Coll. Mus. Comp Zool. The type has very short living chamber, and aperture very similar to that of Mael. praematurum, but wider transversely and with deep ventral sinus, as in some species of Acleistoceras. The general form, aspect, size, and siphon as in Turnus. Naedyceras,^ nobis, includes forms with whorls in section subtriangular, the dorsum broad and flat, the abdomen depressed, and subangular. The siphon is near the venter and nunimu- loidal. The sutures as in Cranoceras, but dorsal lobe more pro¬ nounced, no annular lobes, and no imprassed zone on the dorsum. The genus includes Naed. (Cyrt.) anormale, sp. Barr. pl. 139, (Naut.) vetustum, ibid., pl. 35, in the Silurian, and also a series of degraded arcuate, and gyroceran, Devonian forms which show degeneration in their trochoceran mode of growth. They are also recognized by Professor Hall, as having marks of nautilian affinities. These characteristics could only have been derived from gyroceran ancestors, like Naed. vetustum. Type, Naed. (Troch.) Eugenium, sp. Hall, Nat. Hist., N. Y., Vol. 5,; pt. 2, pi. 58, 59, Mus. Geol. Surv,, Albany. ^ Kpavos, a helmet. 2 NtÍ€\»s, the belly. Hyatt.] 282 [April 4. Oncoceratidae. This family includes forms with peculiarly attenuated apices, or young, the whorl increasing in the adult stages very rapidly in size. The living chambers are generally short, and constricted above, but the apertures are open. Siphon is near the venter, and generally, there is a ventral lobe. The young have straight sutures and these are retained in some adults. Eremoceras,^ nobis, includes arcuate Silurian species with open apertures, short living chambers, whorl in section elliptical ; sutures with dorsal saddles, almost straight lateral sutures and ventral lobes. Type, Erem. (Cyrt.) Syphax, sp. Bill. Pal. Foss., Vol. 1, p. 194, fig. 178, Mus. Geol. Surv. Can. Clinoceras, Mascke, Zeit, d. Deutsch. Geol. Gesell., Vol. 28, p. 49, pi. 1, includes species similar to Oncoceras but with si[)hon between the centre, and dorsum; living chamber and aperture similar, but not swollen, and form very slightly arcuate ; sutures with a minute annular lobe according to Mascke. Clin, (Cyr), exiguum, sp. BUI. Can. Nat. Vol. 5, p. 172, and One. mumiaforme Whit. Geol. Wise.,Vol. 4, 1873-79, pi. 7, have straight sutures and are closely allied to type of Clinoceras in form. Oncocsras, Hall, Nat. Hist. N. Y., Vol. 1, p. 197 is similar in form to (Uinoceras, but the whorl is more depressed, and the ventral aspect fusiform. It has sutures similar to Eremoceras, but the ventral lobe is often narrow, and pointed. The lateral sutures are deeper, and in some species there are dorsal lobes- Siphon is ventral. The living chamber is constricted near the aperture, and much dilated below. One. (Cyrt.) heteroclitum. Barr. Syst. Sil., pl. 118, 475, has the peculiar swollen living chamber of Oncoceras, and ventral siphon. _ Am. Mus. N. Y. Ilercoceratidae» This family has, in the normal forms, trapezoidal whorls with abdomen broader than venter, one line of large tubercles along the edge of abdomen, and sutures with ventral and dorsal lobes. Shells not costated. Ptyssocoras,^ nobis, includes Silurian arcuate shells with 1 "EpT^iioSj alone. ^ XlTuVira), fold. 1883.] 283 [Hyatt. single row of large lateral tubercles, sutures nearly straight; siphon ventral, whorl in section depressed elliptical. Type, Ptyss. i'Cyrt.) alienum, sp. Barr. Syst. Sil., pl. 127, Hercoceras, Barr. Syst. Sil,, Vol. 2, Text 1, p. 152, includes Silurian gyroceran forms, the type of which is Here, mhaim Barr.; but we also include in the same genus all the allied gyroceran, and trochoceran species, like Here. (Gyr.) alatum, sp, Barr. pl. 44, and Here. (Tro.) flexum, sp. Barr. pl. 44, all of which have similar striae of growth, sutures, and ventral siphon. Anomalocerasd nobis, includes but one Silurian species with a nautilian shell, Anom, (Naut.) anomalus, sp. Barr. pl. 34, which has a nautilian form, more involute than in Hercoceras, with smooth, much depressed whorls ; siphon and sutures as in Her¬ coceras. The one-sided position of the siphon is found also in other cases among Tetrabranchiata, and is not probably a generic distinction. Temnocheilus, McCoy, Syn. Carb, Foss. Irel. p. 20. Type, T, coronatus McCoy, j:l. 4, fig. 15. Cryptoceras D'Orb, Prod, de Pal. Vol, 1, p. 58, has for type. Tern, (Naut.) subtuberculatus, Sandb. Verst. Nass. pi. 12, fig. 3, and is a synonym. It includes all the forms with smooth nautilian shells, trapezoidal whorls in section, the venter very broad, the sides divergent, the dorsum narrow and having always an impressed zone. A row of large nodes occurs along the junction of the sides and abdomen. The sutures have broad, ventral, lateral, and dorsal lobes. There are no annular lobes in the Devonian forms, but they ap})ear in some Carboniferous species, as in Tem. latus, De Kon. Cale. Carb. The siphon is ventral in the Devonian forms, but near the centre in most of the Carboniferous species, Centroceras,^ nobis, includes a series of Devonian species with much compressed whorls, abdomen often hollow, sometimes nar¬ row, with one row of tubercles along the edge of the abdomen on either side. The sutures have deep V shaped ventral lobes, deep lateral, and dorsal lobes ; no annular lobes in species observed. The dorsum is frequently gibbous, and has an impressed zone only in the more compressed, and more involute species. Cent, (Cyrt.) ^ Avc'fioXos, anomalous. ^ KiVTpov, a spur. Hyatt.] 284 [April 4, tetragonum, D'Arch. et Vern., Trans. GeoL Soc. Lond. Vol. 6, pi. 30, has young which are identical with adults of Temnocheilus, and adults similar to those of typical Centroceras, but less com¬ pressed. Type, Cent. (Disc.) Marcellense sp. Hall, Nat. Hist. N. Y., Vol. 5, pt. 2, pi. 65, 109. Mus. Geol. Surv., Albany. Hutoceratidae. Species in this family have exceedingly rough shells. The pro¬ jecting lips of the apertures are more or less permanent and often form ridges, or lines of projecting spines or nodes. These may be indefinite in number, but there is a general tendency to reduce them to three rows on either side, and, if carried farther, to one line along the edge of the venter. Sutures have saddles on the venter. Siphon is ventral, or near venter. Zittelloceras,^ nobis, includes species of arcuate Silurian and Devonian longicones with whorl in section elliptical and an exter¬ nal frilled layer resembling Dawsonoceras, but no costae, and much larger ventral sinus in the aperture, and corresponding deflections of the frilled ridges and lines of growth- Siphon is small, tubular, and ventral. Sutures have ventral saddles, lateral lobes, and dorsal saddles. The living chambers are long, and apertures open. Type, Zitt. (Cyrt.) lamellosum, sp. Hall, Nat. Hist. N. Y., Vol. 1, pi. 41. Amer. Mus., N. Y. Halloceras,^ nobis, is confined to the Devonian. The shells have similar frilled layer, flaring lips to the apertures, etc., as in Zittelloceras, but the forms are Gyroceran, or nautilian. The whorl is subtriangular in section. The abdomen is broad, the sides divergent, and the dorsum forms the narrow apex of the section. Thick costae, or rather large nodes are formed along the angles of the sides in the adults. The sutures have ventral and lateral lobes, and in the impressed zone, when this occurs, there may be a corresponding shallow dorsal lobe. The siphon is small, and near the venter. The young are identical with the adults of Zittelloceras. Type, Halloceras (Gyr.) undulatum sp. Hall, Nat. Hist., N. Y., Vol. 5, pt. 2, pi. 53, 54. Mus. Geol. Surv. Albany. 1 Dedicated to Prof. Karl Zittel of Munich. 2 Dedicated to Prof. James Hall of Albany. 1883.] 285 • [Hyatt RutOcerasS nobis, includes arcuate, Devonian forms closely allied to Zittelloceras in the imbricated structures, and flaring apertures of the shells, but having three rows of large nodes on either side. The siphon is ventral, but it is large and nummu- loidal instead of being tubular, and small. The living chambers are shorter than in Zittelloceras, and the form in section is de¬ pressed elliptical. Type is Rut. (Cyrt.) Jason sp. Hall, Nat. Hist. N. Y., Vol. 5, pt. 2, pi. 50, and supp. pi. 124. This genus also includes shells similar to the above, but having the gyroceran form, and less rugged sixrfaces. The abdomen becomes more elevated and slightly narrower, the dorsum slightly flatter, and broader than the venter. Triplooceras^ nobis, includes the remarkable Silurian species Tri. (Naut.) insperatum sp. Barr. Syst. Sil., pl. 461, which has whorls in section like those of Rutoceras, and similar sutures, but with very slight ventral lobe, and the typical three lines of tuber¬ cles on either side ; siphon between centre and venter. Adelphoceras, Barr, ibid. Text 3, p. 788, et suppl. pi. 459, also Silurian, has three lines of tubercles on either side, and though the aperture is contracted as in Gomphoceras, we include it provisionally in this series. Siphon ventral. Kophinoceras,^ nobis, includes Devonian species, which have rough shells as in Rutoceras, and numerous ridges on the abdo¬ men, more or less roughened or broken in adults by nodes formed by the permanent lips of the apertures. The species vary greatly, but are all probably gyroceran, and the normal forms have either two rows of tubercles along the middle of the venter, or a broad raised band as in some Halloceratites. There are also two rows of nodes along the angle of the junction of the sides and abdomen, which are large and persistent. The form in section is more depressed than in Rutoceras, the siphon nummuloidal, and ventral ; sutures with ventral saddles, and in one species a small annular lobe. The type, Koph. (Cyrt.) ornatum, sp. D'Arch. et Vern. Trans. Geol. Soc. Bond., Vol. 6, pi. 28, Mus. Comp. Zool., Camb. has three persistent rows of nodes open to the front, or 1 PvlsT, a fold. 2 TpiirXoos, three-fold. ^ Ko(l>tvos, a basket. Hyatt.] 286 [April 4, spout-like as in the tubercles of Rutoceras. The remarkable species MÛth form and tubercles like Temnocheilus, but a ridged abdomen, the Koph. (Naut.) Coxanum, M. et. W. Geol. 111. Vol. 5, pi. 23, is probably in this genus. Strophiceras,^ nobis, includes a Devonian gyroceran form Str. (Gyr.) binodosum, sp. Sand. Verst. Nass. pi. 12, which has a com¬ pressed whorl with gibbous tuberculated abdomen, and flattened sides. There are several rows of tubercles upon the abdomen as in Kojiliinoceras, but also a central row of tubercles. Sutures with ventral, lateral, and dorsal lobes, but no annular lobes. Form probably close coiled. Siphon yenti*al. The form is peculiar, the young unknown, and we refer the species to this series with great doubt. Solenoceras,^ nobis, includes species of the Carboniferous, Dyas, and Trias having remarkably heavy looking quadrate whorls with furrowed abdomens. Sutures, with broad ventral, lateral, and dorsal lobes, and in adults a small annular lobe may aj^pear in some species. Siphons central and nummuloidal. Liv¬ ing chamber is one-fourth to one-half a volution in length, aper¬ tures with very deep ventral sinus. The shells are smooth except in the young and some adults of Solen, (Xaut.) nodosum, which has a single outer row of large nodes along the sides. Type, Solen. (Xaut.) canaliculatum, sp. Owen, Geol. Ken. Vol. 3, pi. 10. Mus. Comp. Zool. This genus ^ includes a series of Dyassic species described by Waagen Pal. Ind., ser. 13, no. 1. Solen. (Xaut.) transitoriuum sp. Waag, ibid, pi. 6, fig. 4. They have quadragonal whorls, are costated, and have dej)ressed or furrowed abdomens. Siphons below the centre slightly or central ; Sutures with ventral, lateral, and dorsal lobes. Annular lobes? Phloioceras,® nobis, includes the so-called Trematodisci of the Trias described by Mojsisovics in his "Mediterr. Trias Provinz." The shells are ridged longitudinally, and the ridges in the type roughened by transverse striae. Mojsisovics considers them as allies of Xaut. cariniferus, and he may be right, but we have placed them in this series on accomnt of the resemblance of geminatus to Kophinoceras. The sutures have simple, lateral, 1 Srpo€Îov, a twisted rope. 2 SwXtÍv, a groove. ® ^XoLoS) bark. 1883.] 287 [Hyatt. ventral, and dorsal lobes, with small annular lobes. Siphon is central or below the centre. Phi. (Naut.) gemmatum, sp. Mojsis. Das Gehrig. Hallst, pt. 1, pi. 3. Pleuronautilus, Mojsis. Mediten-. Trias Prov. p. 278, in¬ cludes a series of forms with more or less tuberculated, and costated whorls, with sutures similar to the preceding. The siphon is also below the centre. The young of one species, Pleu. subgemmatum, as figured by Mojsisovics, ibid,, pi. 85, is similar to the adult of Phi. gemmatum, and appears to settle the question of affinity. The sutures according to Mojsisovics vary from those having ventral lobes to some having straight ventral sutures in O O O aged specimens. They have annular lobes. Eadoceratidae, This family includes forms in which the whorl in transverse section is some modification of the fusiform outline. The abdo¬ men may be flattened, but is never hollow. The siphons are ventral, or between the venter and the centre. There is a constant tendency to reduce the breadth of the dorsum, and increase the venter. No sulcations or ridges are developed in any genus. Hudoceras, Hall, Nat. Hist. N. Y., Yol. 5, pt. 2, supp. pi. 117, includes straight shells of the Silurian and Devonian, with whorls flattened, sides angular. The sutures have broad, ventral, and dorsal lobes, the ventral lobe deepest; and lateral saddles, which are angular when the sides are angular, and more or less rounded when the sides are rounded. The ventral and dorsal sides of the whorl are equally convex, whorl in section being fusiform. Mus. Geol. Surv. Albany. Tripteroceras,^ nobis, has similar forms and sutures to the preceding, but the lateral saddles are acute. The venter is flat¬ tened, and broader than the dorsum, which forms the a2)ex of the subtriangular section. The siphon is ventral, and nummuloidal, and the whorl arcuate in the young, though straight in the full grown, and the aspect altogether distinct from the shells of the preceding genus. The young are similar to the adults of Eudoc- eras. Silurian and Devonian. Trip. (^Orth.) hastatum, sp. Bill., Hep. Prog. Geol. Surv. Can. 1853-56, p. 333, Mus. Geol. Surv. Can. 1 Tptimip, a rubbing tool. Hyatt.] 288 [April 4, fjdaphoceras,^ nobis, includes species with the young arcuate until a late stage of growth, with whorls fusiform in section, and sutures with dorsal, and ventral lobes, and angular lateral saddles as in adults of Eudoceras, but the siphon shifts from the venter, where it is in the larva, to near the centre. Unfortunately the only figured s])ecies is the Edaph. (Tem.) niotense, M. et W. GeoL 111. Vol. 5, pi. 19. This is selected as the type because it is figured, though in our opinion it is not a full grown shell but only the later adolescent stage of a species as yet undescribed in the collection of Mus. Comp. Zool. The adult in this is close coiled, with flattened sides, broad lateral saddles. An impressed zone appears on the dorsum due to close coiling altering the form in section from fusiform to kidney-shaped, and a V-shaped annular lobe appears in the middle of the broad dorsal lobe. Endolobus, M. et W. Geol. 111., Vol. 2, pi. 25, p. 307. End (Naut.) Avonensis, sp. Dawson, has all the young stages like the preceding genus, but inherits the annular lobes, shifts the position of the siphon to near the centre, acquires the impressed zone, and changes the form of the whorl to the kidney shape at an earliet stage of the growth. The young have broad ventral lobes, bur the adults develop saddles in the centre of these, and in the type, End., spectabilis, large folds or tubercles appear on the sides. The latest survivor of this series is the End. (Naut.) exca- vatum sp. D. Orb. of the Jura, Terr. Jurass. Ceph. pL 80. G on i 0 ceratida e. We interpolate this extraordinary group here because its near¬ est affines are the compressed, straight cones with equal dorsal and ventral sides. This relationship is fully appreciated by Pro¬ fessor Ilall who considers his genus Eudoceras as very closely allied to Gonioceras. Gonioceras, Ilall, Nat, Hist. N. Y. Vol. 1, p. 51 has a broad winged shell, which in form and structure, as indicated by the septa and striae of growth, closely resembles the internal shells of Sepia. We think, the facts are sufficient to warrant our as¬ suming this, as probably one of the passage forms from the compressed Orthoceratites, above described, to the true Sepioidea, and possibly a more or less remote ally of Paleoteuthis Dunensis Roera. of the Devonian. 1 *'E8a4>os, a seat. 1883.] 289 [Hyatt Apsxäoceratidae, The whorls in section are some form of the sub-triangular, and throughout all the genera there is a constant tendency to sulcate the abdomen and retain the gibbous character of the dorsum. The genera have transverse costae, but no longitudinal ridges. The marked characteristic of the group is, however, the persis¬ tence of the large dorsal saddles, which enable us to see, that the dorsal lobe is produced only in the impressed dorsal zone and in this group appears to be due solely to involution. Even when it is replaced by a dorsal lobe in the nautilian species, this dorsal lobe becomes, as in Epliippioceras. subdivided by a small dorsal saddle. Tripleuroceras,^ nobis, includes straight cones, whorl in section triangular, or if elliptical, having flattened abdomen. Siphon is near the venter, nummuloidal, large and usually, but not invari¬ ably, partly filled with deposits more or less radiatory in struc¬ ture. Sutures have ventral lobe on the flat abdomen, saddles at the angles, lateral lobes, and dorsal saddles or comparatively straight sutures around the dorsum. One species, Trip, explora^ tor, sp. Bill., has an additional pair of lateral saddles. Silurian and Devonian. Type, Trip. (Orth.) Archiaci, sp. Barr. Syst. Sil., Vol. '2, pl. 251, 480. The dorsal saddles separate this genus from the Eudoceran group and show them to be the radicals of the following series. Apsidoceras,^ nobis, includes loosely coiled, smooth, costated or tuberculated gyroceran shells, with flattened abdomens. The whorls in section are triangular, the dorsum forming the internal apex of the outline ; siphons near the venter and nummuloidal. The sutures have broad ventral lobes, saddles at the lateral angles, broad lobes on the sides, and doi'sal saddles. There is frequently a line of heavy tubercles on each of the lateral angles of the whorls. They are all large shells and the abdomen is fre¬ quently hollow or fluted along the centre. They occur from Silu¬ rian to the Carboniferous inclusive. Type, Aps. (Lit.) magnifi- cum, sp. Bill. Geol. Surv. Can. Rep. 1853-56, p. 307. Mus. Geol. Surv. Can. Titanoceras,^ nobis, includes Silurian and Carboniferous nau- 1 TpCirXivpos, three sided. 2 the felloe of a wheel. PROCEBDIKGS E. 8. N- H. VOL. XXll 19 JXNUAUT, 1SS4 Hyatt.] 290 April 4, tilian shells similar to the above, but of larger size with a narrow impressed zone on the dorsum, and a corresponding undivided, narrow, dorsal lobe. The sutures have similar ventral, and lateral lobes, but there are a pair of slight lateral saddles near the shoulders. The whorl in section has a narrower abdomen than in Apsidoceras, and longer abdomino-dorsal diameter, and is more compressed or shield-shaped, rather than depressed or triangular in section. Type, Titan.' (Naut.) ponderosura, sp. White, IT, S. Geol. Surv. Final. Rep. on Nebr. Hayden, p. 236, pi. 3. Nat. Mus. iBphippioceras,^ nobis, includes Carboniferous forms with sub¬ acute prominent ventral saddles, broad lateral lobes, sub-acute lateral saddles near the shoulders, and broad, shallow dorsal lobes. In the American species, and perhaps in all, there is a slight dorsal saddle in the centre of this lobe. The septa in all species are creased, or raised into a median ridge between the two saddles. The aperture of Ephip. clitellarium reminds us of Pteronautilus in its shallow, acute ventral sinus. Type is E])hip. ferratum, sp. Owen, Geol. Kent., Vol. 3, p. 574, pi. 10, fig. 2, a species closely allied to Ephip. (Naut.) bilobatum Sow., De Kon. Cale. Carb., pl. 9 but is less involute. The siphon is below the centre in the late adolescent stages, and above the centre, or central in adults of the type species. Pteronautilus, Meek, Pal. Up. Missouri, Smith. Contr., Vol. 14, p. 64, includes but one Dyassic species. This has completely involute whorls, and an aperture extended laterally into wings. The ventral sinus of the aperture is singularly acute resembling in this respect that of Ephip. clittellarium. This slight indication of affinity enables us to ])lace the genus provisionally in this series. Type, Pter. (Naut.) Sebachianus, sp. Gein. Dyas, p. 43, pi. 11. Trxgonoceratidae, The adults of the radical species, and early stages of descend¬ ent forms have whorls similar to those of the Apsidoccratidae. There are, however, longitudinal ridges along the edges of the sulcated abdomens in the adults of the radicals, and in the young of descendent forms these are repeated and then folloAved in Titolv, Titan. 'ECinrioV| a saddle. 1883.] 291 [Hyatt adults by quadragonal whorls. The dorsal saddles are retained, and though dorsal lobes are formed in the naiitilian species, these are often subdivided by a minute saddle. Siphon is above the centre. Trigonoceras,^ McCoy, Carb. Foss. Ireland, 1844, p. 9, Nautil- oceras D'Orb., Prod. Pal. p. 110, is a synonym. Trig. (Gyr.) para- doxiciim, DeKon. Anim. Foss., and Gyr, aigoceras, ibid., D'Orbigny's type are the same species, the latter being the young of the former. The young have lateral costae until a late stage of growth. The abdomen is hollow, the junction with the sides angular, the sides themselves gibbous, and the whorl in section consequently shield-shaped, the dorsum forming the acute apex. The adults retain the form, but lose the costae. The siphon is above the centre. The sutures have broad dorsal lobes, saddles at the angles of the abdomen and sides, lateral lobes and dorsal saddles. This genus seems to be directly transitional to Apsido- ceras, but we have not yet seen the young. The similarities of , this genus and the hollow abdomened forms of the Tribolocerati- dae have been frequently noticed by authors, but we do not regard them as indicating a close genetic connection.^ Stroboceras, nobis, includes Carboniferous species which are similar to Trigonoceras in their larvae, but elevate the abdomen develop two pairs of lateral ridges, and have gibbous inner umbilical shoulders so that the dorsum becomes broader than the abdomen, and decidedly gibbous. The sutures have broad abdominal sad¬ dles, small acute saddles at the lateral ridges, narrow lobes on either side, broad lateral saddles on the swollen or gibbous part of the whorl, and small, sub-acute, dorsal lobes. Apertures are contracted laterally, and dumb-bell shaped. Siphon is half way from the centre to the venter. Type, Strob. (Discites) Hartii. sp. Daws. Acad. Geol. Ed. 3, p. 311, fig. 125, Mus. McGill College. Trematodiscus,^ Meek, was finally established by this author in his Invert. Pal. U. S. Geol. Surv. Hayden, Vol. 9, p. 491, with Trem. (Naut.) stygialis of the Carboniferous, DeKon. Anim. Foss., pi. 45, fig. 11 as th'^ type. The larvae are at firsit stage observed ^ Srpoßos,a vortex. 2 See p. 293, Triboloceratidae. 8 We propose to change this name to Trematoceras since it was used by Häckel in 1860 for the Eadiolaria, and by Ëichwald for Bactrites. Hyatt.] 292 [April 4, similar in form to the larva of Trem. subsiilcatum, and later take on the ridges and abdomen as in Strob. Hartii, but develop a broad furrow between the two ventral ridges. This furrow Avith the tAvo lateral furroAVS form a trisulcated abdomen, and together with the ventral lobes in tlie sutures enable us to separate the species from the adults of Strob. Hartii. The siphon is near the a-enter The small dorsal lobe is first formed in the centre of the dorsal saddle, simple as in Stroboceras. Then a minute saddle arises dividing it into two A^-shaped lobes. This genus also includes sj)ecies with young more evidently similar to the adults of Tii- gonoceras, as is shown in Gaudry's figure of Trem. (Xaut.) sub¬ siilcatum, Ench. du Monde Anim. Foss. Prim. p. 174, but aa'hich sjieedily in course of groAvth ele\\ate the abdomen and deA-elop ridges. In successh'e stages the ridges disappear, and the abdo¬ men and sides become flattened, forming a whorl which in section is typically tetragonal and similar to Discitoceras. The siphon is near the venter in the young, and shifts to midway between the centre and the venter in the full grown. Discitoceras, nobis, is equivalent to the genus Discites, McCoy, Synojt. Carb. Foss. Ireland, p. 17. It includes species with quadragonal aa'horls having the abdomen slightly conA^ex, sides flattened, the dorsum very gibbous. There is also a slight imjiressed, dorsal zone. The young are ridged longitudinally with prominent transverse striae, but though these cross and roughen the ridges, they do not render them subspinous. The sutures have ventral, and lateral lobes, and broad dorsal saddles with small annular lobes. The siphon is aboA'e the centre. The living chambers a-ary from one-fourth to three-fourths of a a^oIu- tion in length. The aperture has a A^ery deep ventral sinus, with large lateral saddles near the dorsum, and small lateral sinuses The type. Dis. costellatum, McCoy, Op. Cit. pl. 2, flg. 4, Avas appar¬ ently the young of a species similar to his Dis. discors. The name Discites has been used by Dellaan, Walch and Schlotheim for genera of Mollusca, and we, therefore, substitute another name for that first announced by McCoy. Phacoceras,^ nobis, has aa'horls compressed and acute in the adults, but Avith young similar to the adults of Discitoceras. The ^^Kos f a leatU. 1883.] 293 [Hyatl whorls are very involute^ and there is a deep, impressed zone of involution on the dorsum. The young sutures are probably similar in outline to those of Discitoceras, but in the adults there are ventral saddles, according to DeKoninck. Type, Phac. (Naut.) oxystomum, sp, DeKon. Calc. Carb. pL 17. Aphelaeceras,^ nobis, includes Carboniferous species allied to Discitoceras until a late stage, but the whorls are more compressed laterally, have hollow abdomens in later stages and adults, and sides more convergent. The forms are gyroceran, and have no impressed zone on the dorsum, which is gibbous, and sometimes projecting along the centre. Sutures, and living chambers and apertures similar to those of Discitoceras. The young appear to have median dorsal saddles, which become divided by slight dorsal lobes during growth. In the adults there is a dorsal lobe, but the median saddle appears to have been absent in the adolescent stage of the single species we have examined. This genus also includes nautilian species, which differ from the typical Aphelaeceras in be¬ ing involute and in having an impressed zone on the dorsum, but the dorsal lobe similar. Aph. (Naut.) difficile sp. DeKon. Calc. Carb. and disciforme, sp. M. et W., Geol. Ill,, Vol. 5, pL 18, are members of this subdivision. Subclymenia, D'Orb, Prod, de Pal., Vol. 1, p. 114, differs from Discitoceras in the sutures, and position of the sqffion. The sutures have a deep V-shaped ventral, and acute, linguiform first pair of saddles, first pair of lateral lobes narrow, a second pair of small, lateral saddles near the umbilical shoulders, and dorsal saddles, divided by shallow annular lobes with a minute median saddle. The abdomens are hollow and the dorsal region gibbous, as in the adults of Aphelaeceras. The siphon is near the vmnter, but the funnels do not approach near enough to inter¬ rupt tlie sutures, or affect the depth of the ventral lobes. But one Carboniferous species is known, Subcly. evoluta, sp. Phil., De Kon. Calc. Carb,, pi. 45. Triboloceratidcie, This family includes shells, which at some stage have longitu¬ dinal ridges rendered subspinous by the transverse striae. The whorls in section tend to become depressed, and in the higher ^ A+eXrs, smooth. Hyatt.] 294 [April 4, species have fluted, and often hollow abdomens. The siphon is above the centre in all except radical forms. The sutures ac¬ quire ventral, lateral, and dorsal lobes, and annular lobes in the higher nautilian species of Vestinautilus. The radical of this family is Thoracoceras among the Orthoceratidae, and this genus could be very appropriately included in this family. The type is the Thor, Vestitura Eichw. Bull, Soc. Imp. de Mose. 1844, p. 761, pl. 17. It includes Thor. (Cyrt.) corbulum, sp. Barr. pl. 125 of the Silurian, and several Devonian species, besides the Carboniferous spinous ridged species like Thor. (Cyrt.) canalicu- latum, sp. DeKon. Calc. Carb. pi. 33, and also the frequently smooth ridged arcuate forms, like Thor. (Cyrt.) Puzosianum, sp. DeKoninck, which have similar transverse striae though less prominent. The connection between this genus, and Triboloceras is too close to need discussion. A connection with the Trigon- oceratidae can also be inferred from the resemblance of the adults of such species as Thor. Puzosianum, and the young of some forms of Trematoceras and Discitoceras. But the connection with Trigonoceras is made very doubtful by the transverse costae of that genus, the form, and the modifications of the dorsal sutures. We incline, therefore, to separate the gi*oup at least provisionally from the Trigonoceratidae. Triboloceras,^ nobis, includes the remarkable series of gyro- ceran, Carboniferous species described by De Köninck in his Calc, Carb., which have subspinous ridges in the young until a late stage of growth, and otherwise resemble the adults of Thor¬ acoceras. The whorl in section is more or less depressed, and either biangular with convex abdomen^and gibbous dorsum, or approximately triangular with concave abdomen. The siphon is above the centre. The sutures have broad, ventral, and lateral lobes, and dorsal saddles without annular lobes. Type, Tribo. (Gyr.) serratum, sp. DeKonnick, Calc. Carb. pi. 32, fig, 5. Mus. Comp. Zool. The forms range from Tribo. (Gyr.) pro- pinquum, ibid., pl. 33, to Tribo. (Kaut.) Meyerianum, ibid., pl. 29. Vestinautilus, Ryckholt, includes nautilian species of the Carboniferous with depressed subtriangular, or trapezoidal whorls, the abdomens very broad, and the dorsum with an impressed 1 TpißoXos, a burr 1883.] 295 [Hyatt. zone in the adults of most species. The young in the adolescent stages have a hollow abdomen with keels, or longitudinal ridges on the abdomen, and forms as in the Triboloceras, but the full grown shells usually become convex on the abdomen with fewer ridges, and in old age are rounded and smooth. Vest, multicari- natus, DeKon. Calc. Carb. pi, 30, remains until late in the adoles¬ cent stage similar to the adult of Trib. Meyerianum. The genus also contains more involute species. In these the carinations tend to disappear, and the whorls become rounded as in Vestin. glo- batus and Coyanus, DeKon. Calc. Carb. pi. 81. The development is much concentrated in the last, the ridges being suppressed at an early stage. Vest. (Kaut,) Koninckii, DeKon. ibid, p. 189, pi. 30, is cited as the type of Ryckholt's genus. Koninckioceras,^ nobis, includes nautilian Carboniferous spe¬ cies with whorls, having a depressed but broad convex abdomen, trapezoidal in section in the adolescent stages, and similar in form to some species of Triboloceras until a late stage of growth. Sutures have slight lobes on the venter or straight, and broad dorsal lobes, but no annular lobes. There is an impressed zone on the dorsum, but the umbilical perforation is very large. Type, Kon. (Kaut.) ingens sp. De Kon. Calc. Carbon, pl. 28, Mus. Comp. Zool. Carnb. Kon. (Kaut.) implicatum, ibid, pl. 18, shows the ado¬ lescent stages. The form of whorl and dorsal lobe appears to place the species in the same series with Triboloceras. Aipoceratidae. This family is remarkable for the rotund form of the adoles¬ cent and adult whorls, and most species have a trumpet-like or flaring aperture. The peculiar heavy ridge of the umbilical shoulders in the nautilian forms is also a marked peculiarity. The siphon is in most forms close to the venter, but in some be¬ tween the centre and the venter. The Sactoceran peculiarities of the radical species separate the group from any series to which it might have been otherwise referred. Aploceras D'Orb. Prod. Pal. Vol. 1, p. 112, includes a series of arcuate Carboniferous forms described by DeKoninck, Calc. Carb. Tliese have a brevicone aspect in most species, the shell ^ Dedicated to Prof. L. DeKoninck. Hyatt.] 296 [April 4, is striated longitudinally with fine closely set ridges even in full- grown shells. The sijjhcn is nuinmuloidal, and above the centre. The whorls vary in section from rounded to elliptical, and even depressed elliptical outlines. The living chamber in Apl. (Cyrt.) rostratum, sp. DeKoninck, ibid, pi. 35, is contracted slightly at the aperture. The type is Apl. Verneuillianum. sp. DeKon., Calc. Carb., pi. 84. The nearest affines of this genus seem to be in the genus Sactoceras. Aipoceras,^ nobis, includes the Carboniferous gyrocei-an forms of which we know but one species, Aip. (Gyr.) gibberosum, de¬ scribed by DeKoninck, Calc. Carb. pi. 32. This has a similar whorl and is evidently a close affine of Aploceras, but the siphon is tubular, and close to the A^enter as in Asymptoceras, The smoothness of the shell, also, is transitional to this last named genus, as well as the more compressed outline of the whorl in section. The sutures retain the simple outlines of the arcuate radicals, having slight ventral and dorsal saddles, or nearly straight outlines. Mus. Comp. Zool. Asymptoceras, Ryckholt, Not. sur Asymp. et Vest. 1852, has for its type, according to DeKon., Calc. Carb. p. 112, Phillips' spe¬ cies of Naut. cyclostomus. Solenocheilus, Meek, is a synonym, having for its type Asympt. (Naut.) Springeri W. & St. J., Tran. Chic. Acad. Vol. 1, p. 124. The whorls increase very rapidly in size, the living chambers are short with flaring, or slightly con¬ tracted apertures. The venter is flattened, or slightly hollow along the centre. The sides are more or less gibbous, and the umbilical shoulders project in heavy ridges, or a large pair of tubercles. Upon each side of these are flutes which are specially characteristic. The dorsum is also remarkable for having the centre gibbous as in gyroceran forms, indicating the recent der¬ ivation of the genus from more loosely coiled forms. The sutures have broad ventral lobes, saddles at the abdominal ridges, broad lateral lobes, saddles at the umbilical shoulders and dorsal lobes, with small annular lobes. Siphon is near the venter. The elliptical form of the young whorl, the large umbilical per¬ foration, the simple, fine, smooth, longitudinal ridges, evenly dis¬ tributed around the whorl, indicate derivation from Aipoceras. Alirtjs. lofty or high. 1883.] 297 [Hyatt The presence of a pair of large tubercles on the chambers of hab¬ itation in some of the species unites them MÙth such forms as Asypm. (Naut.) bifrons sp. DeKon, Calc. Carb. pi. 16. Even the contracted chamber of this species and of Asypm. (Naut.) con- spicuum ibid, pL 19, does not enable us to separate these species.^ Nautilidae. The species of this family have the typical nautilian whorls, ridged in the young, but smooth usually in adults. The sutures, though lobed on the venter in one genus, have generally broad saddles and sutures like those of the recent Nautilus. In several series we traced the appearance of the annular lobe in the sut¬ ures, and an internal septal de¡)ression, which we have called the cone. The sutural lobe appears in the Devonian, and is inherited in the Triassic genus, Cenoceras, with the cone better developed, and becoming separated from the sutural lobe. In different gen¬ era springing from this common type there is a tendency towards concentration in the young, the cones being confined to the earlier stages. Thus in the Jura the cones are still occasionally found in adults, but in the Cretaceous probably very rarely, if at all, and in the Tertiary and present no case of this kind was ob¬ served, though they are characteristic of the young. The lobes in the sutures are not exclusively confined to the larval and earlier adolescent stages of growth of recent species, as are the cones, but may be present in adults of the existing Nautilus. In con¬ sequence, however, of their separation from the cones, they be¬ come easily obliterated in fossils, and are apt to escape observa¬ tion.^ 1 Asympt. (Naut.) dorsale is usually considered the type of D'Orbigny's genus Cryptoceras, Prod, de Pal. p. 114, but the species first mentioned by that author on p. 68, Tern, subtuberculatus, should be considered as the type. 2 This is also the history of the same part in other series besides this family and is applicable to all the families of Nautiloids in which the annular lobes appear. The annular lobe of the suture, and the accompanying depression of the septa are almost invariably united in the Paleozoic forms, and the appearance of the cone is verj'rare. The separation of the two is, however, the rule in the Mesozoic, as detailed above in the Nuutilidae. In the Ammonoidea, on the contrary, the lobes and cones appear though very rarely, in Silurian species, and are fully developed ir the Devonian Gon- latites. Hyatt.] 298 [April 4, Sphyradoceras,^ nobis, includes a series of Silurian and De¬ vonian, annulated, costated, and longitudinally ridged species whose close affinity to, and probable derivation from, Spyroceras, will hardly be disputed. Ridges and costae are both present in the young, but in succeeding stages one or the other, or both may disapjiear. The species are more or less trochoceran in mode of growth, with arcuate and gyroceran forms.^ They have straighter sutures than in Hercoceras besides the differences of the shell markings, and the position of the siphon varies from near the centre to near the venter. The sutures have either straight sut¬ ures, or saddles on the venter, and also saddles gn the dorsum. In some species there is a lateral line of tubercles similar to those of Heroceras, but smaller; and the inner lip of the aperture may also occasionally bend upwards as in Hercoceras. Notwithstand¬ ing these peculiarities the flattened sides and abdomen of adults and their apertures, which are similar to those of Barrandeoceras ; and the compressed elliptical whorls of the young of arcuate spe¬ cies like Sphy. (Troch.) debile sp. Barr. pl. 18, and their central siphon, leads one to associate this genus in the same general series with Baivandeoceras. This group may possibly help us to explain the presence of the longitudinal ridges and annular costae in Barrandeoceras, when more perfect records are available. They may be at present considered a series of shells with slight trocho¬ ceran deformation, which is an offshoot of Spyroceras, and pos¬ sibly nearly related to the ancestral forms of the ÎNautilidae, Type, Sphy. (Troch.) Clio. sp. Hall, Nat. Hist. N. Y., Yol. 5, pt, 2, pi. 59, 111. Mus. Geol. Surv. Albany. Uranoceras, nobis, includes arcuate and g}TOceran forms of the Silurian with large, stout, elliptical, or laterally compressed whorls, abdomen and dorsum convex and about equal in breadth- Siphon is near but above the centre. The sutures have very broad ventral saddles, slight, lateral lobes, and broad dorsal sad¬ dles in the early adolescent stages, and acquire very slight ventral and dorsal lobes in the adults of the American species. The si¬ phon is large and numrauloidal, and the peculiar gibbous whorls, ^ ^ 2i|>upas. a cake of dung. 3 One species, Sphy. (Troch.) nodosum, sp. Barr. pl. 20, fig. 20, appears in only one variety to have had a very faint impressed zone, and may possibly come within our deânition of the nautUian shell. 1883.] 299 [Hyatt and sutures can be more or less closely compared with the young of Barrandeoceras, the adults of Nephriticeras, and the adolescent of the Ifirge Carboniferous forms, like Naut. exirnius, and praegra- vis, DeKon., Cale. Carb. Type, Uran. (Cyrt.) Uranum, sp. Barr, pl. 196.^ I have met with several specimens of this, or a closely allied species, from Anticosti, Mus. McGill. Coll., Bost. Soc. Nat. Hist., and Mus. Can. Geol. Surv. The exact radicals of this genus are unknown, but it has close relations with Spyroceras in its sutures ; and the form of the whorl in U. Uranum is very similar to the laterally compressed and flat abdomened forms of Spyro¬ ceras. Barrandeoceras,^nobis, includes gyroceran and nautilian shells with very large umbilical perforations, and compressed, slightly costated or smooth whorls, generally without an impressed zone, though this is sometimes present. The venter is narrower than the dorsum, the siphon near but above the centre, septa deeply concave, and sutures with ventral saddles, lateral lobes and dor¬ sal saddles, without annular lobes. Type, Barr. (Naut.) natator, sp. Bill. Can. Nat. n. s. Vol. 4, Mus. Geol. Surv. Can., The genus also includes the Bohemian forms Barr. (Naut.) Bohemicum, sp. Barr. Vol. 2, Syst. Sil. pl. 32, 38, Sternberg!, ibid, pl. 36, 37, ty- rannus, ibid., pl. 38, Sacheri, ibid, pi. 39. Living chamber is about one-half of a volution in length ; it is about three-fourths of a volution in length in the type species. Pselioceras,® nobis, includes the series of Dyassic Ophionei, traced by "Waagen in his fossils of the Salt Range, Pal. Ind, Ser. 13, pt. 1, Pisces and Cephalopoda. The large umbilical perfora¬ tion of the type, Psel. (Naut.) ophioneum, sp. Waagen, ibid, pi. 5, fig. 2, shows that it must have been cyrtoceran in the earlier stages for a prolonged period. There is a slight dorsal impressed zone formed, according to Waagen's figures, after the close coiled stage begins, on the first whorl. The resemblance of this genus to the adults of Barrandeoceras natator is very close in the sut¬ ures, and form of the larval and adolescent stages. We place it provisionally in the same series, noting, however, that the forms 1 Barrande has also noted the Nautilius-like aspect of this species. 2 This genus was dedicated to M. Joachim Barrande, before his death, as a token of respect and admiration for .his work upon the fossil Cephalopods. • ^fXiov, a bracelet. Hyatt.] 800 [April 4, also resemble the genus Discitoceras in general aspect during the later adolescent and adult stages. Nephriticeras,^ nobis, includes Devonian forms with ellipti¬ cal or broad kidney-shaj)ed whorls. There is an impressed zone on the dorsum at a late stage in the elliptical forms, and in the kid¬ ney-shaped whorls this appears at earlier stages. The umbilical jierforations are large, and the whorls arcuate until a compara¬ tively late stage of growth in all except the highest smooth forms. Siphon is numrauloidal, and near, but above the centre. The sutures have broad ventral saddles in the adolescent usually with sliofht ventral lobes in the adults. The lateral lobes are also sliirht O O and broad, the doi*sal lobes are similar, and have large V-shaped annular lobes in the impressed zone, the last often being very large, and cones are partially formed. The living chambers are from one fourth to half a volution in length and very broad, the increase of the whorl by growth being very rapid. The longitu¬ dinal ridges in the radical species, such as Neph. (Naut.) comu- lum, sp. Hall, and their arcuate later larval stages, and adolescent gyroceran forms without impressed zones, and central nummu- loidal siphons and sutures lead to the conclusion, that the genus s])rang from some unknown ridged and probably Devonian form similar to Uranoceras, and Aploceras. Type, Neph. (Xaut.) buci- nurn, sp. Hall, Nat. Hist. N. Y., Vol. 5, pt. 2, pL 60, 109. Mus. Geol. Surv. Albany. Cenoceras,^ nobis, includes forms appearing in the Trias, Jura, and Cretaceous, with nummuloidal sijîhons, flattened abdo¬ mens, sides but slightly convex and convergent, the abdomen narrower than the dorsum. The si])hon is nummuloidal, and, though near the centre, may be either above or below it. This again in the young occupies a position on the ventral side of the centre in the few forms known. The shells are nautilian from an early stage with decided impressed zones in the adults. The septal cone becomes in this genus separable from the sutural lobe with which it was combined in Nephriticeras. The transitional Triassic species, like Cen. (Naut.) carolinum, and Tintoretti sp. Mojsis. Med. Trias Prov., may have either ventral saddles or lobes in adults, and siphon central or below centre except in the young, and whorls which are not very involute. Jurassic species, though ^ Nepi'niS) kidney shaped. ^ Kaivds» modern. X883,] 801 [Hyatt with saddles in the young, have almost universally ventral lobes in adults siphon above centre and more involute whorls. The dorsal cones are In some s[)ecies confined to the larval and adoles¬ cent stages, disappearing in adults, whether tlie annular sutural lobes disap}>ear in adults is difficult to determine in fossils. Type, Cen. (Naut.) intermedium, sp. Sow. D'Orb. Terr. Jurass. Ceph. pi. 27. The resemblance of the Triassic forms to Barrandeoceras is decisive for association in the same series, but there is as yet no evidence that Cenoceras was directly derived from Barrande¬ oceras. The frequent presence of longitudinal ridges in the young and in some adults, the appearance of the annular lobes at comparatively early stages, and the kidney-shaped whorls of the larvae, also indicate derivation from Nephriticeras rather than Barrandeoceras. We have, however, not yet found the inter¬ mediate Carboniferous forms, and these, it must be remembered, may prove to be gyroceran and arcuate shells belonging to the common trunk or stock, but not necessarily to either of the gen¬ era mentioned. Some Jurassic species have ridges and transverse striae in the adult shells exactly as in the young of the existing Nautilus. Cymatoceras,^ nobis, includes Cretaceous species of the Ra- diati, remarkable for their transverse costae. The abdomens are rounded and the sides gibbous, though the whorls become com¬ pressed in adults of some species. The sutures have large ventral saddles, shallow lateral and dorsal lobes. The siphons are usually sub-central. Annular lobes and cones are present in the larval and adolescent stages, but disappear in adults. The young, as noted first by D'Orbigny, are devoid of costae, these appearing on the second, or even third whorl in some species. The ventral sutures are distinctly lobed in the later larval stages, the saddles developed later in the adolescent and adult stages. Doubtless at still earlier stages than those observed by us, the suture pre¬ sented the usual larval ventral saddles. Type, Cym. (Naut.) pseudo-elegans, sp. D'Orb. Terr. Cretac. Ceph. pi. 8. Nautilus,^ includes forms of the Jura, Cretaceous, Tertiary 1 Kvjia, a wave. ¿ Reexaminations of the young of this genus have satisfied the author of errors in former views with regard to the funnels. The funnel of the second septum extends to the opening of the coecum of the first septum on the ventral, but not on the dorsal Hyatt.] 802 [Aprü 4, and Present, which have tubular siphons, rounded abdomens pre¬ cisely as in the more radical or lower species of Cymatoceras, and are separable from this genus only by the absence of tranverse costae at all stages. The young sutures in some Cretaceous species, Naut. Dekayi, for example, have first ventral saddles, then in later stages ventral lobes, and in the adults ventral saddles are developed. We have, also, not yet seen a full grown shell among J urassic species, which had the annular lobes though they are some¬ times retained until the latest adolescent stages. The abdomen is O sometimes flattened, as in Cenoceras, but the young even in such species have gibbous, or kidney-shaped whorls, they also invariably present longitudinal ridges, which in some species, may even be said to be slightly subspinous. The sip on in the young is between the centre and the ventral side, and in the Jurassic species retains this position in the adults. In Cretaceous species it appears nearer the centre. In the Tertiary species there seems to have been no lobe developed on the venter at any stage, and in the recent species, all three of which we have examined, there is, also, no abdominal lobe at any stage. The adult lobed stage of the Juras¬ sic progenitors is skipped in accordance with the law of concen¬ tration of development in descendent forms, and the larval saddles are perpetuated without change throughout life. The young of the modern Nautili do not resemble the adults of Cymatoceras or of Cenoceras, and in our opinion have only one nearly related form, the ridged adults of Nephriticeras. We have to go back to the later larval and adolescent stages of the transition forms of the Jura in order to sae that there may have been direct connec¬ tion between the genera Cenoceras and Nautilus, and some of the facts indicate that the separation of the two genera may have taken place in the Tidas. Thus, Naut. striatus, sp D'Orb. and in- flatus, ibid, Terr. Jurass. both have the typical fox-ra of this genus and ventral saddles. Though the siphons are above the centre, and the annular lobes are retained until a late stage, as in Ceno¬ ceras of the same period, they do not resemble them so closely in their aspect as they do the lower forms of the Triassic Ceno- side. The funnel at this stage is, therefore, macrochoanitic, but not holochoanoiinion, that the ventral lobe is due to the approximation of the funnels to the side in Bactrites, and the facts support this conclusion. The Silurian forms we have mentioned are evidently the radicals of Bactrites, and we can only account for their affinities in form, in the transverse striae, and sutures with Mim. compressum by referring both to the com¬ mon stock and imagining them as the more direct descendants Ö O of that stock, which must have had banded shells similar to those of Geisonoceras. The aspect of the apex in two specimens of Bactrites Hyatti, in the Museum of Munich, show that Bactrites is probably a true Nautiloid in the earliest stages, but this dis¬ tinction has lost much of its value since the discovery of the protoconch in some of the straiglit cones. It leaves us still at liberty to give due weight to the transitional characteristics of the later stages. And we must, also, in this connection call atten¬ tion to the important transitional character of the cicatrix, which is shown in our figure of Bactrites on M. Barrande's pi. 490. The cicatrix is unusually large and indicates a protoconch with a larger neck than is usual among Nautiloids. "We have seen some, perhaps, even larger in proportion, but they are exceptional as M. Barrande shows in his figures. These facts show, that the ven¬ tral lobe of Bactrites is an independent production in this series of straight cones. At the same time we should not forget that such parallelisms occur only in nearly related series, which are descended from the same common stock. At least this has been our experience in all the series we have traced, and we are, there- 1883.] 305 [Hyatt. fore, disposed to translate this case in the same way. The artifi¬ cial line between Ammonoidea and îiautiloidea can be placed between this group and Gsisonoceras, with which the shells and characteristics of the transitional forms blend, or between Bac- trites and Mimoceras according to the weight one is disposed to give to the characteristics of the apex. AMMONOIDEA. We have already noted the characteristics of this order, but we find it necessary to add the following remarks. In each group of Goniatitinae there are forms, with whorls depressed and with outlines semilunar in section, which are simi¬ lar to Parodiceras and Anarcestes and resemble these genera also in having undivided ventral lobes. There are also in these radi¬ cal genera and in other genera, either in the adults or in the larvae, forms having simple sutural outlines like those of Parodi¬ ceras with slight rounded saddles, and broad, shallow, angular lobes. There can be but little risk of error, therefore, in assum¬ ing that the genus Parodiceras is the organic centre of the Devon¬ ian forms, and the immediate radical of all Goniatitinae except the Primordialidae. The Primordialidae appear to be directly derived from Nautilinidae, and we have not been able to trace any descendants. This result shows that the genus Anarcestes is probably the genetic centre of distribution, the stock form in the phylum of the Ammonoidea, or the first with undeniably Ammonoidean form and characteristics, the immediate radical, while, as we have noted above, the group of Bactrites and Gei- sonoceras are the distal radicals. Dr. Branco has worked out very valuable results with regard to the law of concentration of development among Cephalopods, but has failed to appreciate their meaning or to note the fact that others had tried to explain similar results. He confirms views first published by the author, that the larvae of the Goniatites and Ammonites have simple sutures, and the ventral lobes undivided in the early stages, and that these lobes be¬ come divided only in later stages, and farther, that the more ancient forms remain longer in the simple or goniatitic stage with undivided ventrals, than the later forms. He alludes also (Pal- PROCEBDIIfOS B. S. V. H. VOU XXtL 20 JAJTUABT, 1S84. Hyatt.] 306 April 4, eontogr. Vol. 3, 1880-81, p. 29) to the Avell known and frequently described increase of complication in the sutures which takes place during growth in every individual, and then shows that the earlier or goniatitic stages are abbreviated in the young of the mesozoic Amraonitinae. He is thus able to compare this process of the abbreviation in later species of the earlier ancestral char¬ acters and correlate its progress with the similar progressive complication, which takes place in the adult fonns in time. He then inquires if this progress in the sutures in adults is an acci¬ dent, " ein reiner Zufall," or if it stands in definite or causal hereditary relation to the concentration or shortening of the simpler or Goniatitic stages of the sutures in the more recent forms. We think that it Avoiild have been no injury to his thorough and remarkable erabryological contributions, if he had noted more fully our remarks on the nautilian character of the first sèptuin in the Goniatites and Ammonites, and of the additional fact that every Goniatite passed through a nautilinian stage, and every Ammonite through a goniatitic stage. Dr. Branco was perhaps misled by our figure (Embryol. of Cephalopods, pi. 3, fig. 3) which has an accidental ventral fissure in the first septum. We distinctly state on page 86, however, that this was due to the "violent removal of the shell," and describe the first septum as having an entire ventral saddle in both Goniatites and Ammo¬ nites on pages 61, 64. These and other facts such as the earlier inheritance of siphonal coecum in the protoconch of the Ammo- noidea, which Dr. Branco erroneously rejects, are due to the law of concentration and acceleration in development, or the shorten¬ ing of the earlier stages of development in more complicated and later occuring species. The same author's supposition that the degenerate forms cannot be distinguished as such, but may be cited as facts against this theory shows that he has not under¬ stood the close coiled embryos of these forms, and is either not acquainted with the researches of Quenstedt, or has failed to consider tlien worthy of his attention. When an author takes high ground in favor of any special method one has a right to expect unusual freedom from error. Dr. Branco's method, how¬ ever, has not enabled him to escape the usual fate of authors, whether fact-worshippers or theorists. He has misinterpreted 1883.] 307 [Hyatt. the most important fact of the embryology of the Nautiloids, the cicatrix, and failed in correlating the resemblances between the embryos of Nautiloids and Ammonoids, and in correctly distin¬ guishing their differences. We appreciate his work very highly, and we gladly acknowl¬ edge discoveries properly belonging to him, and these are, the distinctions of the Asellate, Latisellate, and Angustisellate types of embryo among Ammonoids. The progression of the embryo through the concentrated and accelerated inheritance of ancestral characters was previously stated by ourselves, and his remarks on this point are simply confirmatory, though more fully stated, more completely worked out, and more richly illustrated. Goniatitinae. This sub-order may be characterized by the possesion of smooth shells marked principally by transverse striae, and open aper¬ tures similar to those of Nautiloidea. The large ventral sinuses of the apertures of pearly all the species show, that they must have possesed powerful hyponoraes, or fleshy ambulatory funnels, and were animals which had similar habits to those of Nautiloids, capable of crawling efficiently on the bottom or of swimming by the use of the hyponome. The sutures are entire, witli a few notable exceptions, in the later forms. The funnels change from the macrochoanitic form to a short transitional form, and in some species in later formations acquire the true ammonitoid collar, or in other words, become completely cloiochoanitic. The ventral lobes are undivided, and similar to simple funnel lobes only in the Nautilinidae and Magnosellaridae. In the remaining families genera may have undivided or divided ven- trals in the same family, or exclusively divided ventrals as in the Primordialidae. The dorsal or inner sides are occupied by sad¬ dles only in the gyroceran form, Mimoceras, and the uncoiled or gyroceran larvae of higher forms. The Nautilinidae have a broad dorsal lobe in the impressed zone, but no annular lobes except in one genus, Agoniatites. All the remaining genera have annular lobes, so far as known, except the transitional species from Anarcestes to the Magnosellaridae included in the genus Parod- iceras. The annular lobes are undivided so far as known, the true siphonal saddles having no corresponding dorsal saddles as Hyatt.] 308 [April 4, among the Ammonitinae, Silurian forms are notable as having sutures with undivided ventrals of the simplest type, shallow primitive outlines to the first pair of saddles, primitive first lat¬ eral lobes and dorsal sutures with broad lobes or saddles, but no annular lobes, and the funnels are machrochoanitic. There are exceptional genera, but these are very rare. Devonian forms have the larcre maijnosellanan lateral saddles, and transitional O O ' characters in the forms, septa, and outlines of the sutures, which are advances towards the Ammonitinae. The ventral lobes when divided are apt to be very broad and the siphonal saddles large. In the Upper Devonian and Carboniferous there is a tendency to narrow the ventral lobe and produce smaller siphonal saddles, and divide the first pair of saddles and also the magnosellarian sad¬ dles into smaller lobes and saddles. Forms having marginal lobes and saddles, which make near approaches to the Ammonitinae occur only in the Carboniferous. The abrupt appearance of the Goniatitinae in the later Devo¬ nian and earlier Carboniferous formations of North America, and the absence of radical forms of the group of Nautilinidae in the Silurian indicate that they were migrants from European sources and not autocthonous. 0 Nautilinidae} This family includes forms with gyroceran and nautilian whorls, and sutures with simple lateral lobes either throughout life or until a late stage of growth. There is a dorsal lobe in the impressed zone, which is due to the involution of the dorsal saddle suture, which is present in the larva. There is no annular lobe except in the more aberrant forms. The ventral lobe is a true funnel lobe in the lower species, becoming a wider ventral lobe in the higher species, as in Pinnacites. The funnels are long and tapering, and in species with approximate septa seem to be continuous, though really ellipochoanoidal. The shells are banded with transverse striae, but otherwise smooth, and the apertures have a ventral sinus, which may be in exceptional transi- tiimal forms like Mimoceras quite small, but are usually large. 1 On p. 256 mention of the Magnosellaridae as having undivided ventrals and oi the fact that the doi*saI saddles were confined to the lower forms of Nautilinidae was acci- deutallv omitted. 1883.] 809 [Hyatt The young are asellate, the first suture having no ventral saddle, according to Branco, The larvae in some forms are cylindrical and open whorled, but in all the higher forms closed, and, though still asellate, have the broad form of the ammonitic embryo. Mimoceras ^ includes two well known species of true Goniatites, but these are separable from Bactrites in no essential characteris¬ tic, except the presence of a permanent protoconch upon the apex. The septa have simple concave lateral sutures, and dorsal saddles without annular lobes, the whorls have no impressed zone, and the shells are, therefore, not I'eally truly nautilian in form, but gyroceran ; and even the compressed whorls are similar to those of some Bactrites. The ventral lobe is a simple funnel lobe, as in Bactrites, and it divides the ventral saddles in the same way. Their characteristics, and the protoconch ally them, however, even more closely with Anarcestes and oblige us to place them in the Nautilinidae. This evidence appears to need but one more link^ the finding of a Bactrites with a globular protoconch. Type, Mimoc. (Gon.) compressum, sp. Beyr. Sandb. Verst. Nass. pL 11, fig. 4; also, Mim. (Gon.) ambigena, sp. Barr. Syst. Sil. pl. 3, 12, and possibly Gon. Dannenberg!, Beyr. Verst. d. Bay. Rheins, Ueberg. pi. 1, fig. 5. Anarcestes, Mojsis. Mediterr. Trias Prov. p. 181, was pointed out by that author but insufficiently defined, and a list of species given in a note, the only characteristic cited being the living chambers, which are said to be long. The genus is also character- izable by the broad semilunar whorl, the abdomen broader than the dorsum, this peculiar form is present in the later larval stages, and is maintained even in excess in some very involute species in which the abdomen in consequence becomes excessively broad, the sides very narrow, and the umbilicus very deep. There are some discoidal species, like Anar. crispus S]^. Barr. pl. 9, fig. 81, with rounded whorls until a late stage of growth, but most of the species depart from the tubular outline at a very early age. Sandberger shows that in one variety of Anar. subnautilinus the first whorl is gyroceran, and Branco, in the same species, demonstrates, that variety vittiger is close coiled. Compressed whorls occur in some Silurian species, such as Anar. neglectus, ^ Mtjios, a mimic. Hyatt.] 310 [April 4, sp. Barr., but the compressed forms are more numerous in the Devonian than in the Silurian, like Anar. circumflexifer, and sub- nautilinus Sandb. Verst. Nass. p\. 11. The closest affinités exist between the young of Anar. subnautilinus, and the adults of Mimoceras, and it seems probable that this genus is directly de¬ rived from the gyrocean Mimoceratites, whether through the in¬ termedium of Silurian or protozoic, transitional forms, we do not yet know. Heminautilinus, nobis, includes species with whorls similar to those of Anarcestes, but with angular lateral lobes in the adults. The species figured by Munster as Gon. hybridus, Ueber Clym. pi. 3, fig. 6, is the type. This figure enables us to see that the species was a form having the sutures of Mimoceras and Anarces¬ tes until far advanced in the adolescent stages of growth. Agoniatites, Meek, Paleont. Explor. 40th, Paral. Vol. 4, p. 99, equals Aphyllites, Mojsis. Med. Trias Prov. p. 181. The last author designates the living chambers as shorter than in Anarces¬ tes. Agon VanuxemijSp. Hall, Nat. Hist. N. Y.Vol. 5, pt. 2, pi. 59, has this part of whorl fully two-thirds of a volution in length. Agon, bicaniliculatus, sp. Sandb. Verst. Nass. pi. 11, in the least involute variety fig. 5, has it about the same length, and in the involute variety fig. 5, c-g, it is about one half of a volution. In this last variety the septa are so closely set, that the funnels of the siphon appear to be holochoanoidal, but close inspection brings out the thin connective wall of the siphon which distin¬ guishes the ellipochoanoidal siphon of these groups. This con¬ nective wall is hard to demonstrate, both from its closeness to the shell, and the small size of the posterior ends of the funnels. The sutures have deep lateral lobes, and more prominent saddles than in Anarcestes ; internally the septa are deeper than in Anarcestes, and rise to a shallow concavity near the venter. The dorsal lobe is to be found in the zone of impression, but is very shallow has an elevated suture, and is entire, without annular lobes. Comparison between Mimoceras and Anarcestes, and the larva of Agon. Vanuxemi show that the dorsal lobe, as in Nautiloids, is correlative with the closer coiling of the shell, and is really a su¬ tural depression dividing the large dorsal saddles of the larva. The umbilical saddles of the sutures are formed by the increase in breadth of the zone and the accompanying doi'sal lobes, and in- 1883.] 311 [Hyatt. crease in proportion to the deeper involution of the whorls, as in some Nautiloids. Agon, fecundas, sp. Barr. Syst.Sil.pl. 11, is less involute, and shows how closely the larval shells repeat the adult peculiarities of the parent gyroceran form, Mimoceras. In fig. 4, occurs the extraordinary variety with straight larval apex, which leaves us in no doubt, that these shells must have had a re¬ mote ancestor with a straiixht cone in the adult sta<ïe. Airón, fide- O O ^ Iis and verna sp. Barr. pis. 8,9 and Vanuxemi, Hall, exhibit spe¬ cies with the highly concentrated development common in the more involute, and compressed shells which generally terminate the series we have studied. They have skipped the larval pecul¬ iarities of tlie gyroceran stage, and become close coiled and even involute on the first or second whorl. Agon, tabuloides, sp. Barr, pis. 4, 244, exhibits in the section figured a decided annular lobe, showing that this may occur in some Silurian species of the Ammo- noids, though usually a Devonian characteristic of both Ammonoids and Naiitiloids. Another curious fact is that it has the internal depression we have called the cone, which among the Xautiloidea is not in our experience fully developed until after the Carbon¬ iferous, and certainly must have been rare, if it occurred in other forms in the Paleozoic. Pinnacites, Mojsis. Med. Trias Prov. p. 181, was merely mentioned by that author and the type given, as the Pin. eraacia- tus sp. Barr, of the Silurian. We have not examined the type, but in the Mus. Comp. Zool. there exist fine specimens of the De¬ vonian form which does not differ apparently from its Silurian ally. The highly compressed and acute whorl becomes excessively invo¬ lute at an early age, and we were able to follow the sutures far enough to see, that in the larva the natural decrease of involution must bring about the disappearance of the umbilical lobes, and reduce the sutures to the outlines of Agoniatites. The abdomen is also broader in the young and the whorl in section is identical with the adults of Agoniatites, and these facts indicate direct de¬ rivation from the latter. The septa of the adults are double con¬ cave in correspondence with the lateral lobes, the internal surface being divided by ridges corresponding to the lateral saddles. There is a broad dorsal lobe, with two small, widely separated dorsal saddles, the impressed zone is very deep, and there is no annular lobe, unless the broad median lobe may be so considered Hyatt.] 312 [April 4, CelaecerasS nobis, includes only Cel. (Gon.) praematurum, sp. Barr. Syst. Sil. pl. 522. It is unique among Silurian forms in the sutures, which possess outlines similar to those of the more complicated Devonian and Carboniferous forms. There is in Bar- rande's figures, though not described by him, ventral saddles, similar to the saddles of Pinnacites and on either side are two deep lateral lobes similar to those of Glyphioceras. The first pair of lateral saddles are large and hastate, and the second pair of lateral saddles broad and rising rapidly to the umbilicus. The shell is not very involute, showing that it is a member of a larger series, which probably had both less involute and also pos¬ sibly some more involute members. The inner lateral saddles as¬ sociate it with the Agoniatites rather than Anarcestes. We re¬ gard it as having probably the same relations to Agoniatites, that Herainautilinus has to Anarcestes. The young, as in Heminautil- inus, will probably be found to repeat the parent form until a late larval stage in some species, M. Barrande repeatedly alludes to this species as one of his best illustrations of anachronic species, or species which are out of place in time ; which more closely resem¬ ble succeeding forms of more complicated structure than those of the fauna in which they occur. To us they are simply highly specialized forms, which have adopted habits similar to those of the species they resemble, and have been accordingly modified in their adolescent and adult stages, but still retain in their larval and adolescent stages the marks of their recent origin from the lower forms with which they are associated. We have accord¬ ingly named this form, the racer [Kote.] CLYMENINAE This group has characteristics which are so evenly balanced that we should have found it difficult to decide whether it was Nautiloid or Ammonoid, if it had not been for the protoconch, and the young sutures, which are shown by Branco in the Paleontogr. Vol. 3, 1880, pi. 8, fig. 1. The sutures in the young have ventral lobes, and the broad ventral saddles of the group are developed later. It is, therefore, an Ammonoid, but we cannot say that the young resemble the lower Goniatites, since the first suture of the only species known, in place of being asellate, or straight, has a broad saddle as in the higher Goniatites. A still more remarkable peculiarity, if gen¬ eral, is, as stated by Brauco the absence of depressions on either side of the neck of ^ KcXt|Sí a racer. 1883.] 313 [Hyatt Cloiochoanites. [Transitiones.] The transitional forms of the Goniatitinae, which have the short type, or partly cloiochoanitic funnel elevated upon a median the ovisac. These are the remants of the umbilical perforation, which are present in the young of all other close coiled Ammonoids. These charactenstics, and the dor¬ sal position of the siphon, and the presence of deep annular lobes, are differences of great importance and show that we miist place the series above the Nautilinidae. The forms are evidently highly concentrated in development, but descendants probably of the same stock as Anarcestes. The adult forms and the sutures of Cyrtoclymenia are very similar to those of Anarcestes and indicate this derivation. The group appears to have had a very narrow distribution chronologically and geographically, and was probably a highly specialized series with exceptionally rapid evolution in some open fields of the Devonian. So far as I know, not a species of this group has yet been found in North America, those described heretofore are now known to be Goniatites. The author has spent considerable time in the study of this group and divided them into genera, but these can only now serve as the basis of appreciative criticism for the elaborate work of Dr. Gumbel, Ueber Clym. Paleontog. Vol. 11, p. 83, 1863. This author's sub-groups are equal to our genera, and most of his varieties are what we should call species. We, therefore, use his names in this value without making any claim to the credit of having originated them. This extraordinary series shows the phenomena of quick evolution in three series of forms. Cyrtoclymenidae with a series beginning with an Anarcestes-like form and passing through discoidal and compressed to quadragonal costated forms. Cymaclymenidae, a similar parallel series but with more complex sutures, and Gonioclymenidae also a similar series, but with more involute forms than the last, and the sutures becoming ammonitic with median ventral lobes and saddles divided by a pair of marginal lobes. The whole range of the transformations of the Goniatitinae are paralleled in this short series, whose principal differential charac¬ teristic is the dorsal position of the siphon. We have had no opportunity of studying the siphon but Dr. Gumbel's gi'oup of Euclymenia with imperfect siphons appear to us like the imperfect siphons which occur not infrequently among Nautiloids, the con¬ nective wall being destroyed by maceration. The Nothoclymeniae are apparently those with longer, larger funnels and thicker connective walls. Dr. Gumbel's figure of Cly. speciosa gives the funnels as if terminated by a darker colored connective wall. CYRTOCLYMENIDAE. • The sutures are simple with broad undivided ventral saddles, rounded or incomplete shallow lateral lobes and only rarely internal saddles on the sides. The siphon is tubu¬ lar and small. Cyrtoclymenia includes species with depressed semilunar whorls in section, similar to those of Anarcestes. The sutures are similar, but the ventral snddlea of the Clymeninae are present, though rounded, and the lateral lobes are also rounded. Type, Cyrt. angustiseptata, Gurab. pi. 15. Oxyclymenia includes forms with discoidal shells, compressed and more or less in¬ volute whorls. The larval depressed whorls of Cyrtoclymenia have disappeared from the adult stages. The sutures have ventral saddles more prominent than in that genus, and in some species the angular lateral lobes are deeper, but there are no large lateral saddles. Type, Oxy. laevigata, ibid, pi. 16, also undulata, pi. 17, and Dunkerl, pi. 16. Hyatt. J 314 [April 4, siphonal saddle are not separable from congeneric forms having undivided ventral lobes. The Primordialidae have very large siphonal saddles carrying short, collarless funnels, and the Magnosellai'idae have undivided ventral lobes. The funnels are small and elevated upon the first pair of saddles in most of the Magnosellaridae and these so closely simulate si¬ phonal saddles as to deceive the most accurate observer. Plalydymenia^ nobis, includes species with siraihir sutures to Cyrtoclymenia, but the whorl is discoidal as in the Oxyclymenia and it differs from both of these in hav¬ ing costated whorls, which are in section sub-quadragonal Platy. annulata, ibid, pi. 15, fig. 11, and includes, also pi. 18, fig. 12, spinosa, pi. 16, fig. 1, and binodosa, pi. 19, fig. 1, though the latter has an internal lateral saddle. CYMACLYMEWIDAE. The forms have undivided ventral sadlles, and two pairs of lateral lobes, the first pair angular and the second pair in the umbilicus and rounded. The siphon is transi¬ tional to the large siphoned species of the next family. Cyinaclyinenifi includes species with forms similar to those of Cyrtoclymenia, but m")re compressed and involute. The sutures have prominent ventral saddles, and two l iteral lobes with two umbilical or dorsal lobes on either side of the annular lobe. Type, Cyra, striata, ibid, pi. 18, and bllobata, pi. 19. These forms in their siphons and their sutures connect Cyrtoclymenia with the larger siphoned groups, the Nothoclymeniae of Gumbel, which begin with the next genus. Selhicymenia includes species with whorls compressed, sub-quadragonal and precisely similar to Platyclymenia but the sutures are similar to those of Cy- maclyrnenia and the siphon also according to Dr. Gumbel is large and complete. Type, Sel. angulosa, pl. 19. GONIOCLYMEXIDAE. The ventral saddles are divided by median lobes, and by a pair of marginal lateral lobes. There are also two lateral lobes on either side. The siphon is large, and the funnels conical. Cryptodymeniaf nobis, includes but one species. This has the form of Cyrtoclymenia and enables us to connect the series of genera having ventral lobes, with this radical genus. The sutures have median lobes dividing the large ventral saddles, and two pairs of lateral lobes. Type, Crypt. Beaumonti ibid, pi. 20, fig. 5. Cydodynenia Includes an extraordinary discoidal species with numerous, slowly growing whorls. The sutures have the ventral saddles divided by median lobes. Goniodyineiiia includes species with quadragonal whorls, sometimes costated as in Scllaclymenia and Oxyclymenia but they have mediau ventral lobes, and two pairs of lateral lobes and large siphons. Type, Gonio. speciosa, ibid, pi. 20, and alsosubarmata and intermedia ibid, pi. 21. The Gon. pessoldes V. Bucü, and biimpressus, ibid, Gon. et Clym. pi. 1, probably are true Glymeninaeas stated by Gumbel, and may be species of this genus. DUcodytnenia includes Clymenia Haueri, which has sutures very similar to those of the preceding, but the lobes are more numerous, and the annular lobes broad. The whorls are much involved and compressed in the adult with a deep impressed zone. 1883.] 315 [Hyatt The similaritv of the funnels and funnel lobes to those of the Nautilinidae, may be seen in Sandberger's figures in Jahrb. d. Nass. Vex\ f. Naturk. Vol. 7, 1851, pL 2. The Glyphioceratidae and Proleceanitidae have undivided ventral lobes in tlie lower or transitional genera, and divided ventral lobes and cloiochoa- nitic funnels in the hierher and later occurrine: forms. Triainoc- O o eras and Pronorites, have transitional ventral lobes which may be considered as undivided, and all the remaining genera have the short, coilarless type of funnel. The excessively short, rapidly narrowing funnels, and the similar character of the breaks which they make in the sutures of the ventral side, is a marked, and highly interesting peculiarity, and led to our designating the whole group at first as Epichoanites.^ "We supposed that all the following genera could be included in this group, and farther characterized as having simple funnels without collars. Dr. Branco, however, in Zeit. Deutsch. GeoL Gesell. 1880, p. 607, note, made us aware that Beyrich's observation on Glyph. (Gon.) sphaericum, and Sandberger's figure of Glyph. (Gon.) crenistria Verst. Nass. pi. 5, fig. 1, showed that the siphonal col¬ lar was present in both of these species. Branco's statement that our definition of a collar may be regarded as a mere question of opinion cannot be sustained. The microscope makes it evident that the fore-reaching part of the septum is not like the posterior part. The fore-j^art is an open collar around the siphon, while the hinder part is a true funnel, interrupting the siphon as in our figure, Embryo. Ceph. Bull. Mus. Comp. Zool. 3, No. 5, pi. 2, fig. 15. Sometimes also as in the cases above cited, and in the tyj>e of Schistoceras described farther on, it is easy to see with unassisted eyes, that the true funnels below and the collars above are distinct from each other- Primordialidae. This group which is closely allied to the Nautilinidae, possese remarkable interest as the most primitive series in which many of the essential differentials of the Ammonitinae first make their ap¬ pearance. The ventral lobes are very large, and are divided by 1 We are not yet satisfied that this would not be a convenient descriptive designation for the whole of the Cloiochoanites [Transitiones] or all the Goniatitinae, except the Nautilinidae. Hyatt. J 316 [April 4, median siphonal saddles which carry the funnel lobes. The de¬ velopment of the sutures has been followed out by Branco, Sand- berger, and the author, and all agree that these median saddles arise in the centre of the primitive ventral lobes. The two arms of the primitive ventral lobe become widely separated, and appear in the later stages of growth, like true lateral lobts. The first pair of saddles are large and rounded, though in the higher and more involute species, often angular. The ridges from these cross the septa and have corresponding dorsal saddles on either side of the annular lobe. The young are assellate, as first stated by Dr. Branco, and the first sutures and radical forms indicate direct derivation from the same stock as Anarcestes. While still in the broad whorled anarcestian stage, the septa are nautiloidian or concave, but when the deep ventral lobes and large lateral saddles are formed, the septa become ammonitoid or convex along the median line. The funnels, which are also anarcestian in the young, become shortened as the siphonal saddles arise and assume the ammonitoidal aspect, though no collars are developed. The funnel lobes are variable in size, and may be absent in some specimens, as in two of Gephuroceras lamed in Coll. Mus. Comp. ZooL This variation has been also observed by Sandberger in this same species and in same variety with one of the above cited instances namely, lamed Var. cordatus, Jahrb. d. Nass. Ver. f. Îî^aturk. Vol. 7,1851, pl. 3, fig. 21, 22. The early larval sutures have broad dorsal lobes, which become narrowed into annular lobes in later stages correllatively with the development of the ammonitoid septa, and especially with the development of the first pair of lateral saddles. We adopt Beyrich's name of Pri- mordialidae as more significant than that subsequently given by Sandberger. G-aphuroceras,^ nobis, includes species with discoidal young, whorls with broad and more or less flattened abdomens in the adolescent stages, and the sides divergent as in Manticoceras. There is a decided resemblance to the adults of this genus espec¬ ially to Man. latidorsatum, but the larval and adolescent stages in the radical species are separable by means of their flattened abdo¬ men and less rapidly growing whorls. The adult whorls become ^ F€4>\ipai & bridge. 1883.] 317 [Hyatt compressed and subacute in several species, but retain their open umbilici except in the most involute species. In these, however, the larvae remain sufficiently constant to enable us to recognize the genus. The serrated or costated abdomens of the young of several forms, is also in marked contrast with Manticoceras. The depth of the ventral lobes causes the septa to assume a convex aspect, but the median line remains concave until a late stage of growth. The large lateral saddles when first formed, and until a comparatively late stage in radical species, have no correspond¬ ing dorsal saddles, these arise later as two minute saddles in the dorsal lobe, on either side of the annular lobe. The species are Geph. (Gon.) calculiforme, Beyr. Sandb. Verst. Nass. pi. 8, fig. 9-9a (not 9c-d), aequabile, Beyr. Gon. Mont. Rhen. pl. 2, fig. 1, and also Sandb. pi. 8, fig. 10, to which last also belong Sandberg- er's figures 9c-d mentioned in brackets above, Hoeninghausi, D'Arch. et Vern. Trans. Geol.Soc. Vol. 6, n. s. pi. 25, and Buchii, ibid, pi. 26, fig. 1, serratum, sp. Sandb. pi, 9, fig. 8, planorbe; sp. Sandb. pi. 9, fig. 3, forcipifer, sp. Sandb. pi. 6, fig. 8, discum, Roem. Nordw. Hartz. Paleontogr. Vol. 6. pi. 13, fig. 35 (not pi. 6, fig. 7) and bisulcatum ibid, pi. 6, fig. 8, acutum. Keysei', Dora. Schief. Verh. Mimneral. Gesell. St. Petersb. 1844, pl. A fig. 6. In America we have seen only Geph. (Gon.) complanatum, sp. Hall, Nat. Hist. N. Y. Vol. 5, pi. 70, fig. 6-12, and the Type, Geph. sinuosum, ibid, pi. 70, fig. 73-75, Mus. Geol. Surv. Albany. Manticoceras,^ nobis, includes species with compressed and often very involute whorls, which are, however, directly traceable by the closest gradations into forms with broad whorls, open um¬ bilici, and an aspect similar to that of Anarcestes. The young are invariably less discoidal than in Ge})huroceras, the abdomens rounded, and the sides divergent outwardly. A close resemblance to Agoniatites bicaniliculatus, or tuberculosocostatus ^ occurs in the costated young and in the sutures and form of Mant. tripai ti- tum until a late larval stage. The adult sutures have the same general aspect as those of Gephuroceras, but the septa in the compressed involute forms become more decidedly convex. The 1 MavTiKosi prophetic. 2 This is a costated species of Af^oniatitites, which we have supposed to be equiva¬ lent to Sandberger's figure Verst. Nass. pi. 2, fig. 3. It also resembles costulatus D'Arch. et Veru. Trans. Geol, Soc., pi- 26, fig- 8. Hyatt] 318 [April 4, lobes remain rounded until later stages of the growth, the funnel lobes are generally smaller, the larger lateral saddles are also more persistent and retain their forms unchanged even in the ex¬ treme old stages of the largest specimens. The species are as follows, Manti. (Gon.) latidorsatum, Sandb. Verst. Nass. pi. 8, fig. 8, trif)artitum, ibid, fig. 7, lamed, fig. 4, 6, are all regarded by Sand )erger as varieties of lamed, but they can be distinguished by their development from each other. Manti. (Gon.) intumescens, Beyr. Rheins. Ueberg. pi. 2, fig. 4, Sandb. ibid, pi. 7, Manti. (Gon.) complanatura, pi. 8, fig. 5, the latter a variety of lamed Sand, both have in adults excessively acute, and involute whorls. Manti. (Gon.) bisulcaturn, sp. Keyser, Dom. Schief, Verh. Min. Gesell. St. Petersb. pl. A, fig. 7, seems to be in this genus. In America we have found, Manti. (Gon.) simulator, sp. Hall, Nat. Hist. N. Y. Vol. 5, pt. 2, pi. 69, fig. 1,2, Pattersoni, ibid, pi. 72, fig. 1-5, Gon. Goniolobus, Meek, Geol. Expl. 40 Parall. Vol. 4, pi. 9, fig. 5, may possbly be a greatly modified Carboniferous form of this genus with closely approximated septa. Type, Manti. simulator, sp. Hall, Mus. Geol. Sur v. Albany. Magnosellaridae, This family is distinguished by the early development and undivided outlines of the magnosellarian saddles, which induced Sandberger to designate the group as the Magnosellares. The ventral lobes are primitive and undivided, and in the lower forms are small, and really mere funnel lobes breaking the outlines of the ventral suture. This is a purely nautilinian character and the first pair of saddles also retain a similar primitive aspect, but in the higher forms tend to split up and form a first pair of lateral lobes and a second pair of saddles. The first pair of saddles have no correspondent on the dorsum but the second pair have corres¬ ponding dorsal saddles when they are present, as in Maenoceras and Sporadoceras. These facts justify the opinions of Mojsisovics, and others that the sutures of this genus are approximate to those of the Nautilini, and our own observations indicate closer affinity for Anarcestes, than for any other genus. The figures given by Sandberger in Jahrb. d. Nass. Ver. f. Naturk. Vol. 7, 1851, pi. 2, of Parodiceras biarcuatum, and amblylobum illustrate this affinity for Anarcestes, especially since these two have 1883.] 319 [Hyatt no annular lobes on the dorsum. The septa are concave as in the Nautillni in the larvae, but become convex internally after the rise of the maornosellarian saddles. The ma2:nosellarian sad- O O dies, besides their prevalence in the Devonian species and in the larvae of later forms, are notable as resembling the similar large saddles of some of the Nautiloids, as in the young of Enclirnato- ceras, and the rise of the septa dorsally in some of the Anarcestes group. Parodiceras,^ nobis, includes species with whorls in section semilunar, sutures with very broad, generally flattened, and primi¬ tive forms af the first pair of saddles, lateral lobes shallow and mostly angular, the raagnosellarian saddles, often rising, as in the young of Toimoceras above the level of the first pair on the venter. Our opportunities for study in this genus have been limited but we have been able to see that some species have very small annular lobes and others, as figured by Sandberger in Jahrb. d. Nass. Ver. Naturk. above quoted, had probably no annular lobes. These characteristics of the septa and the general aspect of the adults are transitional from Anarcestes to Tornoceras. In Branc. umbilicatum according to Sandberger's figure, in Jahrb. Nass. Ver. Vol. 7, 1851, pi. 2, fig. 9, there are slight saddles on the dor¬ sum corresponding to the first pair on the venter, while in biar- cuatum and amblylobum there are none. The first named species is directly transitional from Parodiceras to the genus Brancoceras. The young are stouter than in Torn auris and they also inherit the semilunar and involute whorls at early larval stages, Parod. curvispina is discoidal and biarcuatum, planilobum, an- gulatum are all more involute species figured by Sandb. Verst. Nass. pi. 10, but the last is laterally compressed. Parod. amblylobum, ibid, pi. 10, fig. 8, and pi. 4, fig. 5, has a compressed whorl in the later larval stages. Parod. oxycantha Sandb. pi. 10, fig. 3, in some specimens shows a decided tendency to division of the ventral saddles as in Maenocöras bifer. Parod. (Gon.) sub- lineare Munst. Ueb, Clym. et Gon. pi. 4, fig. 5, according to Mun- ster's figure is transitional to the simpler sutures of Parod. (Gon.) undulosum, sublaevis, and globosum, sp. Munst. Ueber Clym, et Gon. pi. 4. We have found but one species in this country, the 1 HopoSvs. transition. Hyatt.] 320 [April 4f type of our genus, Parod. (Gon.) discoideum, sp. Hall, 13, Rep. on State Cabinet and also Nat. Hist. N. Y. Vol. 5, pt. 2, pL 71, fig. 1-13. Tornoceras,^ nobis, includes species which are similar to Far- odiceras, but have compressed whorls, and annular lobes. The sutures have rounded saddles on the venter and rounded lat¬ eral lobes with the typical magnosellarian saddles of the family. The first pair of saddles have no corresponding saddles on the dorsum, and the annular lobes are situated immediately between the large dorsal saddles corresponding to the magnosellarian saddles. The ventral lobes and sutures in the larval stages are similar to the adults of Anarcestes. The stage at which the ven¬ tral saddles are in a primitive condition has close resemblance to the older stao:es of Parodiceras. Sandberger's figures of Torn. O O O (Gon.) circumflexum, Verst. Nass. pl. 10, fig. 9, and auris, pl. 10 a, fig. 19, and Hall's figure of the young of uniangulare show this very well. The effect of the lateral saddles in elevating the in¬ ternal parts of the septa and the concavity of the outer parts of the same sutures, which remain concave throughout life, are well shown in Sandberger's figures of Torn, auris, pi. 10, a, fig. 4, 5. Torn. (Gon.) discum, sp. Roem. Nordw. Hartz. Paleontogr. Vol. 3-pl. 6, fig. 7, and auris, ibid, pi. 6, fig. 11, both have sutures even in late stages, perhaps adults, very similar to those of Anarcestes, according to Roomer's figures. Torn, auris, sp. Sandb. pi. 10 a, fig. 12, and Var. pi. 10, fig. 11-12, together with Var. pi. 10 a, fig. 18, 19, make together a series which we include under the same name, Var. undulatus, pi. 10, fig. 17-19, also belongs to this species. Torn. (Gon.) retrorsum, sp. Von Buch, has a stout variety similar to auris, the typical variety undulatus Sandb. pi. 10 a, fig. 7, has morecompressed whorls than auris even in larval stages. The young in both of these varieties are smooth, « and exhibit no signs of the raised and costated abdomen of auris, though here and there, specimens with reversionary characteristics occur. Var. typus. Sandb. pi. 10, cannot be separated in the larva or adults from the above. Torn, acutum, sp. Sandb. pi. 10, fig. 10, pi. 10 a, fig. 2, can be separated by the sub-acute whorl of the adolescent stage and the breadth of the later lobes and narrow ^ Topvos* circular or rounded. 1S83.] 321 [Hyatt ventral saddles. In America we find Torn. (Gon.) Mithrax, sp. Hall, Nat. Hist. N. Y., Vol. 5, pt. 2, pi. 69, bicostatum, ibid, pi. 72, fig. 8-10, and the type, Torn (Gon.) uniangulare ibid, pi. 71, 72. Maeneceras,^ nobis, includes forms in which the first pair of saddles are broad, somewhat flattened, and during growth tend to become subdivided. The inner angles become subacute, and small marginal lobes appear between the outer angles and the ventral lobes. These changes may take place late in the life of theani- mal or be inherited at comparatively early stages in accordance with the lawof concentration of development in the more compli¬ cated species. The dorsal sutures change correllatively, and acquire a pair of subacute dorsal saddles on either side of the annular lobe as in Sandbe^ger's Maen. terebratum, Verst. Nass. pi. 5, fig. 3 d. Sandberger's figure 4 of Maen. bifer, and his fig¬ ure 5 c of Maen. (Gon.) delphinum show, that as in Sporadoceras the dorsal saddles were connected with a second pair of saddles, and not probably with the piiraitive saddles. The larval stages, are identical in form and sutures with Parodiceras. Maen. (Gon.) bifer, sp. Sandb. ibid, pi. 9, fig. 4, is closely allied to Parod, oxycantha but has the genetic characters of this genus in the later larval stages. Maen. (Gon.) terebratum, sp. Sandb. ibid, pi. 5, fig. 3, exhibits all these characteristics, and the changes in the first pair of saddles described above at still earlier stages. The adult form of this species is the Maen. (Gon.) acuto-laterale Sandb., pi. 6, fig. lb. The whorl in this stage is no longer rounded, but subacute, and the first pair of saddles are completely divided by rounded lobes. We have studied the intermediate adolescent stages of this species, and have been able to connect the two figures quoted by a fine specimen precisely intermediate in size and all its characteristics. Coll. Mus. Comp. Zool., type Maen. acuto-laterale, sp. Sandb. Sporadoceras,^ nobis, includes species which like Maeneceras have additional saddles developed from the division of the first or primitive pair of saddles, but these are generally pointed in adults and the first and second pairs of lobes hastate. The forms of the shells still remain similar to those of Parodiceras and Maeneceras and the magnosellarian saddles are undivided. A 1 M-rivq, a crescent. a 2hropa8os, scattered. PBOCKKDLNGB B. R. H. H. VOt.. XXIL 21 FEBHCAfiY. ISRA Hyatt.] 322 [April 4, specimen in the Col. Mus. Comp. Zeel, either identical or closely allied to Spor. bidens exhibits primitive saddles incompletely divided although the shell is about 4.3 cent, in diameter. The ventral lobes, and the first pair of marginal lobes are still com¬ paratively small and are elevated upon the large primitive, larval saddles. This genus also as in the preceding shows two dorsal saddles which arise in correlation with the formation of the additional second pair of saddles, and are connected with them by ridges traversing the septa. The first pair of saddles have no correspondents on the dorsum. Sandberger's figure of Spor. bidens, pi. 8, fig. 11, illustrates these observations completely. The annular lobe is deep and undivided. Spor. (Gon.) bidens, sp., Sandb., pi, 8, fig. 11, bilanceolatum, ibid., pl. 9, fig. 7, and Spor. (Gon.) contiguum, sp. Munst. Clym. et. Gon. pi. 3, fig. 8, orbiculare, ibid, pi. 5, fig. 4, cucullatum, sp. Von Buch, Gon. et Clym. pi. 1, fig. 4, Hoeninghausi, sp. Von Buch, lieber Amm. et Gon. and Munsteri, ibid, pi. 2 ; fig. 4, 5, also Sand. pL 5, fig. 2, are examples of this genus. Type, Spor. bidens, sp. Sandb. Glyphioceratidae. This family includes forms having whorls semilunar in section and discoidal shells, with some more compressed and involute forms. The ventral lobes are undivided in the lower transitional forms which connect the family directly with Parodiceras, among the Magnosellaridae, but in the higher genera they are divided by siphonal saddles with cloiochoanitic funnel lobes. The first or primitive saddles are on the venter, the first or primitive lateral lobes are contracted in breadth, deep, and often angular, the magnosellarian saddles are undivided in some genera and divided in others by a single pair of wide, shallow, angular lobes. The primitive, or first pair of saddles have corresponding saddles on the dorsum, on either side of the deep, undivided, annular lobes. The development and outlines of the sutures, and the ventral lobes ex¬ hibit the close affinities of this family for the Magnosellaridae, and the still younger stages, as described by Branco, show their affini¬ ties for the Nautilinidae, especially Anarcestes. The septa are convex along the median line, in all the genera except in the transitional genus Brancoceras. The development of the siphonal saddles is similar to that of the Primordialidae, but 3883.] 323 [Hyatt the sïradations of the forms and sutures are unmistakable and o connect them genetically with the genus Parodiceras. The si- phonal saddles are evidently in this and other families purely representative characteristics arising from the tendency to divide the primitive ventral lobes. Within the families, however, it is evident that the siphonal saddles are inheritable after they have been introduced, and become fixed in the organization. In each of the three sub-families, the siphonal saddles are in accordance with this law first independently generated, and then become fixed. The essentially representative character of the division of the lobes and saddles when first introduced in each series is especially well exhibited in the Diraorphocerae. In this tribe the ordinary outlines of the sutures of this family are modified by the introduc¬ tion of marginal saddles, which sub-divide the lobes, as among the more recent Ammonitinae. After careful investigation, we can find no evidence for the supposition that the recent Ammonitinae ac¬ quired their similar modes of dividing the lobes by direct inheri¬ tance from such highly involute Carboniferous species. On the contrary there is strong evidence that the Triassic Ammonitinae sprang from discoidal shells with forms of whorl more closely resembling the primary radical Anarcestes. If we analyze the forms of Triassic species, as they have been published by Mojsisovics in his classic work Med. Trias., Prov. we are at once struck by the prevalence in every series of a certain proportion of discoidal forms, and by the fact that these are repeated in the young of the more involute forms of the same series, as is not infrequently noted by Mojsisovics himself in his descriptions. Some genera like the Sphingites among Arcestidae and Tirolites, and Xenodiscus may have all, or nearly all the species discoidal, but as a rule the variation from these forms to much more involute, or completely involute forms takes place in the same genus, and is useful only in distinguishing the species. The discoidal or less involute forms are always the simplest in the structure of the sutures, as well as larval in their own series ; witness again Sphingites and Xenodiscus, also such single species as are found in the group of Trach. furcosa, and Mono- phyllites. Arcestes, Pyas, and Trias, so far as we know, is the only group as a whole, which possesses in the adults the depressed larval Hyatt.] 324 [April 4, form of the Silurian Anarcestes, "w hich we also recorded as occur- ing in nearly every group of the Goniatitinae and at the morjdio- logical hase of the grouj s of the Clj nicninae. This form of Avhorl has almost disappeared in the Trias and its ])lace is taken by quad- ragonal derivatives, like Xenodiscus, and the compressed helmet shape, which "we have designated as the secondary larval forms. Whatever form we may admit into the category of secondary lar¬ val radicals, they must in all cases be discoidal, with open umbil¬ ici; and either themselves, or their imimcdiate ancestral forms must be shells, without spines or prcmintnt outgrowths, thoi gh they may be costated, as in Sandbergeoceras, the immediate radL cal of the Prolecnnitidae. The exact agreement between the sutures and forms, and their development botli in the series and in the individual is not in the Trias substantiated by the observed geological positions of many species, in fact the occurence of Monophyllites, Sj)hingites, and Ceratites Sturii and the second¬ ary radical, Dinarites Mohainmedanus and other instances are directly against these views. This, however, does not at all alarm us, if there is any truth in the theory of descent, we feel sure that the clue lies in the development of the individuals which occupy the lower morphological borders of each series, and exhibit in their forms, sutures, and shells, the nearest approach to the primary radicals of tlie Silurian or Cambrian. The Paleozoic or primary radical is Anarcestes, and its depressed whorl becomes larval in the more involute forms of all the genetic series with which we are acquainted among the Paleozoic Gonia¬ titinae, being absent from the early stages only in forms with gyroceran young, and it is inherited by all forms above the Car. boniferous, at the earliest stage of the formation of the apex or conch. The discoidal secondary radicals are unquestionably the nearest allies of this primary radical, -which occur in the Devonian, and these in their turn have asimilar relationship with the Dyassic and Triassic species having similar discoidal forms and simpler sxitures than other more involute shells of their several groups- We can, therefore, with the reservation, that the connections have not been actually made, state that a great change takes place in the Dyas, and that here or about the time of the end of the Paleozoic the secondary larval radicals, or Mezosoic helmet- 1883.] 325 [Hyatt shaped with more complicated lohes and saddles like Monophylli- tes, Ceratites Sturii, Dinarites Mahammedanus, and the less invo- Inte, smooth, similar, forms of Gymnites, Ptychites, perhaps also Pinacoceras, and Ophiceras begin to be more numerous and to replace the Paleozoic radicals as generators of distinct genetic series. Psiloeeras, thoaga in the Jara, belongs, we think to this class of radical forms. The tertiary radical forms of the Jura and Cretaceous are often highly ornamented and complex in their sutures, but they are also discoidal, like all other larval radicals and they give rise to pro¬ gressive series of more involute forms as we have tried to demon¬ strate in several families, and numerous genera. Following out Iläckel's nomenclature which we regard as truthful and express¬ ive, these categories of radicals can be very appropriately desig¬ nated as the Epacmatic, and Acniatic radicals, and when as is often the case, the Acmatic forms become the radicals of degenerative series of uncoiled forms, we propose to call them Paracmatic radicals. [Gastriocerae.] Brancoceras,^ nobis, includes species of the Devonian and Carboniferous with undivided, ventral lobes, rounded or spatulate first pair of saddles, broad, abrupt, magnosellarian saddles, and only one pair of often hastate lateral lobes. The spheroidal form of the shell, the deeply involute, and semilunar whorls show close affinities for Parodiceras. The adult staijes still continue to O 4 have the septa concave along the centre, and only convex in the reorion of the ma2:nosellarian saddles. The annular lobes are largue O O O and undivided, and the first pair of saddles liave corresponding saddles on the dorsum. The dorsal sutures have also larcje O saddles corresponding to the magnosellarian saddles as in the Magnosellaridae and three lobes. The internal sutures are, therefore, precisely similar to those of Sporadoceras, but the external sutures have one pair less of lobes and saddles, and in that genus the first pair of saddles on the venter have no corre¬ spondents on the dorsum. The closest connection evidently ex¬ ists between this genus and Munsteroceras, which have similar outlines in the sutures, and rounded first lateral lobes as in this 1 Dedicated to Dr. W. Branco. Hyatt.] 326 [April 4, genus. Bran. (Gon.) sulcatum, sp. Munst. Ueber Clym. et Gon. pi. 3, fig. 9, ovatum, ibid, pi. 4, fig. 1, lineare, ibid, pi. 5, fig. 1, subsulcatum, ibid, pi. 5, fig. 2, and Ungeri, Munst. Beitr. are all good examples of this genus. The type, Bran. (Gon.) Ixion sp. Hall, Nat. Hist. N. T., Vol. 5, pi. 78, which is a close ally, if not identical, with Bran. (Gon.) rotatorium, DeKon. Calc. Garb, pi. 47, fig. 12, has young, ibid, pi. 73, fig. 12, which are similar in their globular shells, and in their sutures to the adults of the more spherical species of this genus. Branc. umbilicatum, sp. Sandb., has external sutures very similar to those of Parodiceras and is mentioned in the description of that genus. Munsteroceras,^ nobis, includes species of the Upper Devon¬ ian or Lower Carboniferous which resemble Brancoceras closely in form but are more discoidal. The siphonal saddle is small or larval in shape and is developed in the centre of a deep, often straight-sided ventral lobe. The first or primitive saddles are on the venter and spatulate in outline. The magnosellarian saddles are undivided on the sides in the lower species of this group, but a curious and instructive variation takes place in American forms. Munst. parrallelum has the outer suture of the magnosellarian sad¬ dle depressed posteriorly below the level of the inner suture of the corresponding dorsal saddle. An acute lobe is thus formed on the edge of the umbilicus, and a small saddle rises inwardly from this towards the zone of compression, and is visible near the edge of the umbilicus in Munst. Oweni. In Gastrioceras this lobe be¬ comes more important, and in Paralcgoceras it is a lateral lobe, and plainly visible on the sides. In the upper Devonian, accord¬ ing to Hall, we find Munst. parallelum, nobis, equals Oweni var. parallela. Hall, Nat. Hist. N. Y., Vol. 5, pt. 2, fig. 1, 2, pi. 73, the members of this genus are Munst. (Gon.) TVhitei, nobis, equal Oweni, sp. White, Geol. Ind. Ann. Rep. 2, 1880, Pal. pi. 7, fig. 3, 4 (Gon.) tumidura, sp. Roem. Paleontogr. Vol. 3, pi. 13, fig. 3, 4 (Gon.) implicatum, sp. Phill. Geol. York. pt. 2, pi. 19, fig. 24, 25, reticiilatum, ibid, fig. 26-32, excavatum, ibid, fig. 33-35, compla- natum, DeKon. Calc. Carb. pi. 46, fig. 4, inconstans, ibid, pi. 48, fig. 4-9, truncatum, ibid, pi. 48, fig. 2, 3, implicatum, ibid, pi. 50, fig. 2. The forms ai-e all more involute than Munst. Whitei, 1 Dedicated to the memory of Georg, Graf zu Münster. 1883.] 327 [Hyatt. which is a discoidal shell. Type, Munst. parallelum, sp. Hall, Mus. Geol. Surv. Albany. Grastrioceras,^ nobis, includes species with open umbilici and whorls in section semilunar or trapezoidal. The abdomen is wider than the dorsum and the sides often costated. The larvae always have whorls semilunar in section, and similar in form and sutures to the larvae of Brancoceras during the stage before the genera¬ tion of the siphonal saddles. The siphonal saddle is developed as in the Primordialidae, but is not usually so large or broad as de¬ monstrated by Branco and confirmed by our own observations. The sutures at earlier stages and the whorls have the usual simil- o arity to those of the adults of Anarcestes. The adolescent and adult stages have deep, straight-sided ventral lobes and siphonal saddles which are often more or less prominent and angular. The first pair or primitive saddles are on the venter and often spatu- late making the sutures similar to those of Munsteroceras, while the general aspect of the shells is representative of Glyphioceras. The magnosellarian saddles are divided at the shoulders of the whorls with acute lobes, and internally are flanked by saddles ris¬ ing rapidly to the zone of impression as in Munsteroceras. Gast. (Gon.) Listeri, sp. Phill. Geol. York. pi. 20, fig. 1, of the Carbon¬ iferous and Marianus, ibid, pi. 27, fig 2, and Jossae, sp. M. V. K. Russ. and Ural Mts. pi. 26, fig. 2, both from the Dyas are European examples of this genus, and in this country we find Gast. (Gon.) Kingii, sp. H. W., U. S. Geol. Surv. Expl. 40th Parall. Vol.4, pi. 6,fig. 8, and (Gon.) globulosas, sp. M. W., Geol. Surv. Ill, Vol. 2, pi. 30, fig. 2. Paralegoceras,^ nobis, is similar to Gastrioceras in its sutures, but the whorls are compressed, and the sides smooth in the cast. The sutures are similar also to our genus Schistoceras, but the broad, shallow, angular second lobes, and the peculiar internal saddles show closer affinities for Gastrioceras and Munsteroceras. It is apparently a compressed Carboniferous form directly con¬ nected with Gastrioceras. The only species known to us is the Paral. (Gon.) lowense, M. et W. Geol. Ill, Vol, 2, pL 30, fig. 3, from the coal measures. ^ Fao^piS, a pot-belly 2 IXapoXc^ci), I lie by the side of Hyatt.] 328 April 4, [P rioDOcerae.] Prionoceras,^ nobis, includes species with broad, acute, straight- sided, undivided, ventral lobes similar in outline to the first pair of saddles, and the first pair of lobes. The niagnosellarian sad¬ dles are undivided. The young of these shells, if they had the same mode of development as other Goniatites must have had more rounded first pair of saddles at some stage of growth, and this would render the sutures at this stage similar to those of some forms of Parodiceras. The immediate affinities, however, connect them with Brancoceras, from which genus we should not have dared to separate Prionoceras, but for the artificial necessity of showing clearly the genetic relations of this genus and Glyphio- ceras. The species differ from Brancoceras only in the acutencss of the first pairs of saddles and lobes, and their peculiar pyramidal shape. The type is Pri. (Gon.) divisum Munst. Geber Clym. et Gon. pi. 4, fig. 6, of the Devonian. This species and Pri. Belvalia- num, De Kon. Anim. Foss. pi. 49, fig. 5, and Calc. Garb. pi. 50, figs. 8-10, have sutures similar, in their undivided ventral lobes, and lateral outlines, but the first pair of saddles on the venter are pyramidal in shape, like those of Glyphioceras, and are evidently transitions to this last genus. Glyphioceras,^ nobis, includes species with whorls in section semilunar, trapezoidal or compressed, the abdomens broad and convex, the sides divergent outwardly, and frequently costated. The sutures are remarkable for the acute, angular outlines of the lobes and saddles in the adolescent and adult stages, and the large size and frequently bottle-shaped siphonal saddle. There are exceptions to the angularity of the lobes in some spec¬ ies, which retain the early larval form of the outlines of the sut¬ ures in their later stages of growth, but in these, the rapidly nar¬ rowing ventral lobes, and the large size of the linguiform first lateral saddles enables one to refer the species to their proper genus. In their later larval stages the sutures are not distinguish¬ able from those of Munsteroceras, with which also the forms of the whorl agree equally closely. The siphonal saddles are small, and occupy only the apex of the straight-sided, deep ventral lobes, and the first pair of saddles are spatulate, and the lateral lobes and magnosellariau saddles are precisely as in that genus. 1 CLpUv, a ^ rXv4>i&, the notch iu an arrow 1883.] 829 [Hyatt Notwithstanding the extraordinary development of the siphonal saddles, there is no corresponding division in the annular lobe, which is subacute and entire in outline. The first pair of saddles have corresponding but smaller and less acute saddles on the dor¬ sum, and the magnosellarian saddle is also nearly as prominent as on the venter. The umbilical shoulders are occupied by rounded lobes. All the known species are found in the Carboniferous and have the pyramidal form of siphonal saddle. The funnels are surmounted by collars which can be seen on good casts of Glyph, sphaericum and which closely simulate the collars of Schistoceras.^ It seems evident that Prionoceras is the intermedium between this genus and the common family radical, Brancoceras. The young of Glyp, obtusura, as figured by DeKon. Calc. Carb. pi. 47, shows the relation of the group to the more remote radical Par- odiceras with which the sutures agree quite closely except in the siphonal saddle. Glyph. (Gon.) crenistria, sp. Phill. Foss. Corn. Devon, pi. 50, fig. 234, spiralis, Rom. Nordvvest-Harzegeb. Paleontoger. Vol. 8, pi. 8, fig. 15, carbonarium. Buch. Ueber Goniat. pi. 2, fig. 9, sphaeri¬ cum, Calo. Carbon., pi. 47, fig. 8-5, striatum. Sow. Min. Conch., trans., pi. 53, fig. 1, striolatum, Phill. Geol. York. pt. 2, pi. 19, fig. 14-19, truncatum, Rom., Nordwest-Harzegeb. Paleontogr. Vol. 8, pi. 18, fig. 80. The shells in this section of the genus are involute but the umbilici open„ and the whorl semilunar in section. The ventral lobe has a distinct siphonal saddle, which divides it into two acute terminations, the first pair of lateral saddles are inclined towards the umbilicus, and often acute, first pair of lateral lobes also acute, the magnosellarian saddles are undivided. The second section includes the following, Glyph. (Gon.) obtusum, sp. Phill. Geol. York. pt. 2, pi. 19, fig. 10-18, micronotum, ibid, pi. 19, fig. 22,23, platylobum, ibid, pi, 20, fig. 5, 6, stenolobum, ibid, ph 20, fig. 7-9, Barbotanum. M. V. K., Russ. and Ural, pi. 27, fig. 3, diadema, ibid, pi. 27, fig. 1, complicatum, Kon. Calc. Carbon, pi. 50, fig. 4, nitidum Phill. ibid. pi. 20, fig. 10-12, These shells have similar sutures, but the saddles are often slightly rounded or linguiform, and the whorls compressed. The siphonal saddles are smaller, and larval in shape and proportion. All of the first list except spiralis occur in the lowest beds of the Carboniferous, and I See also p. 315, and p. 336. Hyatt,] 330 [April 4, those of the second list in the formations, up to the Lower Coal Measures of Bigsby and Etheridge. [Dimerocerae.] Dimeroceras,^ nobis, includes a single species, Dim. (Gon.) mamilifer, sp. Sandb. Verst. Nass. pi. 5, fig. 5, which has rounded first pair of saddles on the venter, rounded first lateral lobes, and an additional pair of shallow, broad, angular, lateral lobes near the umbilical shoulders, generated by the division of the magnosellarian saddles. The ventral lobes are undivided and this species is evidently a transitional from Brancoceras to Pericyclus. Pericyclus, Mojsis. Mediter. Trias. Prov. p. 141, includes two Carboniferous species which were designated by that author without description. The spatulate first pair of saddles and the additional, broad, angular, lateral lobes, which arise from div¬ ision of the magnosellarian saddle give the sutures exactly the aspect of those of Dimeroceras. The genus can be separated only by the divided ventral lobes, and costations. Type, Peri. (Gon.) princeps, DeKon. Anim. Foss. and Calc. Carbon, pi. 49, fig. 1. [Dimorphocerae.] Homoceras,^ includes only the curious species (Gon.) calyx, sp, Phill. Geo). York. pi. 20, fig. 22,23. This has trapezoidal whorls, like those of Gastrioceras, and sutures wliich are quite sim¬ ilar to those of Glyphioceras. The form is, however, dwarfed with sutures still more like those of Nomismoceras, which has sim¬ ilar dwarfed forms. The siphonal saddles are small, the ventral lobes broad, and open, tlie first saddles and lobes rounded, and the magnosellarian saddles short, undivided, and rounded. The sut¬ ures show the species to be in the same series as Nomismoceras, but the highly depressed, trapezoidal form of the whorl obliges us to separate it from that genus. Gon. mutabilis Phill. ibid, pi. 20, fig. 24, 25, not fig. 26, may also belong to this genus, but there are no sutures as yet known. Nomismoceras,^ nobis, includes a series of dwarfed forms with compressed whorls, open umbilici, resembling in minature the larger forms of the Prolecanites but differing from these 1 in two parts. ^ lQ^O|jiicr|JU)s ^ coin. 2 'On-osdike, 1883.] 331 [Hyatt. decidedly in their sutures. The ventral lobes are divided, and very broad and large, the lateral sutures generally are very like those of the second section of the genus Glyphioceras with which the forms are evidently closely allied, but the small siphonal sad¬ dles, and the larger size, and rounded outlines of the first pair of saddles, and the aspect of the magnosellarian saddles are differen¬ ces of some importance. Nomismoceras (Goniatites) spirorbis Phil. Geol. York., pi. 20, fig. 51-55, rotiformis Phil, ibid, pi. 20, fig. 56-58, paucilobus Phill. ibid, pi. 20, fig. 36-38, platylobus Rom. Nordwest. Harzge. Paleontogr. Vol. 3, pi. 13, fig. 32. The whorls are discoidal, compressed, and helmet shaped in section in the first two species, and involute, compressed, and subtrigonal in the last two. Nom. paucilobus in form and sutures, is intermediate be¬ tween the discoidal forms of this genus and Dimorphoceras, Dimorphoceras,^ nobis, includes Carboniferous species with involute compressed whorls, and sutures quite distinct from those of Nomismoceras on account of their peculiar siphonal saddles, narrow first pair of saddles and divided lobes, but resembling them closely in their magnosellarian saddles, and general aspect. They are in fact, only more complicated and modified examples of the same style of sutures, the lobes having ammonitic marginal sad¬ dles in place of entire outlines. They have a narrow, prominent, siphonal saddle, and minute funnel lobe, the arms of the ventral lobe on either side of this, are divided by one or two minute mar¬ ginal saddles, the first lateral saddle prominent spatulate, the second lateral lobe divided like the arms of the ventral, the magnosellarian saddle broad and undivided. But two species are known to us. Dim. (Gon.) Gilbertsoni, Phil. Geol. York., pi. 20 fig. 27-31, LooneyL Phil., ibid, pL 20, fig. 32-35. Prolecanitidae, This family can be distinguished by the absence of the great magnosellarian saddles, which are so completely divided as to be more or less unnoticeable in adults, though visible as an underly¬ ing outline in some radical species, and in some larval forms such as Sand. Chemungense. The number of lobes and saddles varies greatly, but there are never less than two pairs of large, lateral ^ Ai|j.op4>os> double formed. Hyatt.] 332 [April 4, lobes in adults exclusive of the inner series of auxiliary lobes and saddles. One pair of lateral lobes, the outer pair of the adult stages, arises out of the division of the first pair of saddles in Pro- lecatiites, and several pairs in Beloceras, but whether this may be considered exclusively the method of generation cannot be stated now. The primitive pair of lateral lobes are generally deepest and may be distinguished by this character. The general form of the matured lobes is hastate and the saddles club-shaped, but the Belocerae,the lowest or radical sub-family, and the lower or rad¬ ical species of Triainoceras have simple flexuous outlines in the sutures. The ventral lobes are, as in other families, undivided in radical genera and species ; and divided in higher forms of each tribe or sub-family group. The first pair of saddles have corres¬ ponding saddles on the dorsum, and there are deep, undivided, annular lobes. The dorsal suture is divided not only by this pair of saddles, but has adlitional pairs corresponding closely to the number of saddles and lobes on the exterior in all the broad whorled genera ; except the umbilical lobes, which as in all the Goniatitinae have no correspondents. The radical genera appear to approximate to Sandbei*geoceras and this genus is not trace¬ able to Parodiceras, so far as we know. The other radical forms, however, like Sandbergeoceras Cnemungense and Phar. tridens have lobes and saddles, which indicate derivation from Magnosel- jaridae, and we have accordingly referred the group provisionally to this radical. This family is very interesting also, because we can directly trace a connection with Jurassic forms, and see that the phylliform marginal saddles of Lytoceras and Phylloceras are probably derived through Monophyllites of the Trias from Prole- canitidae. The line is evidently an unbroken one to Ptychites and we think these and the Pliylloceratidae can be distinguished by the excessive division of the lobes, the attenuated and deeply cut saddles and the peculiar form of the siphonal saddles and their close relations to the first pair of saddles taken in connection with the phylliform marginal saddles, and the tendency to produce ad¬ ditional external saddles, and lobes, and numerous auxiliary lobes and saddles. We shall probably in forthcoming publications sep¬ arate this whole series, including Arcestes from the Ammoni- tinae and Goniatitinae, under the name of Prolecanitinae. 1S83.] 833 [Hyntt [Belocerae.] Sandbergeoceras/includes species with peculiar saddles and obes, which are variable in number and larval in their outlines. The shells are discoidal, the form of the whorl in section is de¬ pressed, semilunar, the abdomen broadest, and there are costae from an early stage of growth. The ventral is undivided accord¬ ing to Sandberger's figures, and the funnel lobes break the suture with a peculiar tubular prolongation of the tips of the lobes, and with slight shoulders or minute incipient saddles on either side, like those in the undivided ventrals of Brancoceras. These are transitional to the minute siphonal saddles observed in the ven¬ tral lobes of Triainoceras, present in the lower Devonian forms. The species are Sandb. (Gon.) tuberculosocostatum, sp. Sand. Verst. Kass., pi. 4, fig. 1 ; and Sandb. (Gon.) Chemungense, sp. Hall, Nat. Hist., N. Y., Vol. 5, pt. 2, pis. 69 and 74. The last species has sutures which show it to be closely allied to Prole- canites. Though very simple in outline, they would not have been sufficient to separate it from Prolecanites but for the pres¬ ence of costaticns on the whorls. Beloceras,'^ nobis, includes only one species but the most remarkable and in many characteristics the most instructive of Devonian forms. The additional lobes and saddles are very numerous and have entire margins and are apparently only limited in number by the breadth of the whorl. They are partly derived from division of the ventral or primitive ventral saddles, and partly from the division of the magnosellarian or umbilical saddles. It must also be observed that here as in other forms the primitive pair of lateral lobes are marked by their greater size, and their earlier development. Saudberger has traced the mode of genesis of the lobes and saddles in his text and in figure of Beloc. (Gon.) multilobatum, sp. Beyr., and Branco, Paleontogr., Vol. 27, ser. 3, pl. 6, fig. 6, has shown the process in its earlier stages. His figures and state¬ ments, however, do not make it perfectly clear, that the first pair of saddles are true primitive saddles, though they certainly seem to have a close resemblance to those of'Maenoc. bifer. Mojsiso- vics in his " Gebirge um Hallstatt, p. 48 and 69 refers the type of 1 Perticated to Prof. Guido Sandberger. ^ BcXos, an arrow. Hyatt.] 334 [April 4, this genus to his genus Pinacoceras, from which, however, it is separable by the characteristics given above, and by the very simple flexuous lobes and saddles of the adolescent stage, which resemble those of Sandbergeoceras. [Sagecerae.] Medlicottia, Waagen, Paleontol. Indica., ser. 13, pt. 1, p. 39, and p. 88, equals part of Sageceras as formerly defined by Moj- sisovies. The lobes are divided by single linguiform marginal saddles, and the second species named below has trifoliate, or divided saddles, though the first pair and the marginal lobes have the same form as in Orbignyanum. The ventral lobes are deep and apparently undivided, the first pair of saddles are narrow long, and the margins cut by several lobes and saddles growing progressively longer internally. The numerous auxiliary lobes are generated apparently as in Beloceras from the marginal divisions in the outlines of the first pair of saddles, and from the division of large magnosellarian saddles near the umbilicus. We only know of two species which can be properly included in the genus. They are Medlic. (Gon.) Orbignyianum, sp. M. V. K. Russ and üral Mts., pi. 26, fig. 6, and Medlic. primas, Waagen, Pal, Indica., ser. 13, pt. 1, p. 39 and 83, pl. 2, fig. 7, both Dyassic forms. Mojsisovics has pointed out the probable connection of this group with Beloceras and Sageceras in his Med, Trias, Prov. p. 183, and in this we also concur. There is no proof in support of Branco's opinion, that forms like these with divided lateral lobes can be separated from true Goniatitinae, but the contrary appears to be evident in all forms of the Carboniferous, Dyas, and Trias. Sageceras, Mojsisovics, Mediterr. Trias. Prov. p. 187, Das Gebir. urn Hallst, p. 69, is a Triassic genus, which, as shown by this author's admirable figures, has the closest relations of affinity with both Beloceras and Medlicottia. The lobes are more nu¬ merous than in Medlicottia and are divided by small saddles, which are themselves slightly denticulated, showing a nearer approximation to the marginal saddles and lobes of Ammonitinae than any other Goniatite except Cyclololus. Of the three species mentioned by Mojsisovics, two S. Haidingeri, sp. Hauer, and S. Wal¬ ten, sp. Mojsis. are European. In America Sag. Gabbi, Mojsis., 1883,] 385 [Hyatt equals Sag. Haidingeri, Gabb. Geol, Calif. Pal. Vol. 1, pi. 5, fig. 8-10. This genus is separable from Medlicottia by the flattened outlines of the first pair of saddles, and from Beloceras by its divided lateral lobes. We do not feel sure that the ventral lobes are undivided, though here as in Beloceras the generation of the auxiliary lobes and saddles on the venter is similar to the same process in Maeneceras, which takes place in the outlines of the first pair of saddles. These saddles and the sutures generally are • similar to Beloceras. while the form of the whorl and abdomen show close affinity for Medlicottia. Lobites, Mojsisovics, Das. Gebir. um Hallst, pt. 1, p. 155 and Med. Trias. Pro v. p. 176 are apparently distorted or retrograde forms with sutures similar to those of this family. Whether they are really members of this family and can be traced to an origin in genera like Popanoceras or Sageceras or some allied groups, we have no proper means of ascertaining. Mojsisovics regards them as genetically connected with Maeneceras delphinum in which the living chamber is similar, having a contracted shape which appears at first sight to justify this opinion. We, however, can regard such resemblances as genetically important only when species are similar in the sutures, and also found in closer relations in time. [Prolecanites.] Prolecanites, Mojsisovics, Med. Trias. Prov. p. 199, includes species with more or less discoidal forms, smooth, compressed whorls and a variable number of lobes and saddles. The ventral lobes are undivided. The lobes are hastate and the saddles club- shaped. The first pair of saddles become divided to form an additional pair of lobes and saddles and the auxilliary lobes and saddles are formed by division of the magnosellarian saddles. According to this author the type is Prol. (Gon.) mixolobum, sp. Sandb. Vorst. Nass. pi. 3, fig. 13, not pi. 9, fig. 6, which is a Pro- norites. Besides the species enumerated by Mojsisovics we include in the same genus also Prol. (Gon.) lunulicosta, sp. Sandb. ibid., pl. 3, fig. 14 with five pairs of lateral lobes and saddles. As we have remarked above on p. 333, Sandb, Chemun- gense, sp. Hall, Nat. Hist., N. Y., Vol 5, pt. 2, pi. 69, makes a very near approach to this genus. ^ 1 Lecanites, Moisisovics, Mediter. Trias. Prov. p. 200, which has for its type Lec. glaucus, ibid. pi. 30, is described by the author as genetically connected with Pro- Hyatt.] 336 [April 4, Pharciceras,^ nobis, tyis genus can be readily separated froni Sandbergeoceras which it very closely resembles in form, and in the general aspect of the sutures, by means of the divided ventral lobes and the smoothness of the whorls. Phar. (Gon.) tridens sp. Sandb. Verst. Nass. pi. 4, tig. 2, is discoidal, and (Gon.) clavilobum ibid. pi. 8, fig. 3, is very involute. There are two pairs of lateral lobes in tridens and a small umbilical lobe, and two pairs of dorsal lobes, and in clavilobum Sandberger figures four pairs of lateral lobes. Phar. (Gon.) multiseptatum, as figured by Quenstedt, Die Ceph. pi. 8, fig. 8, is a member of this genus. Schistoceras,^ nobis, includes a single species which is not figured or described, but can be readily distinguished by its large, bottle-shaped, siphonal saddle. This is the only characteristic by which it differs from Prolecanites. The two arms of the ventral lobe are widely separated, and there are only three pairs of lateral lobes and a small umbilical lobe tvith two pairs of dorsal lobes. The lobes are hastate, and the saddles more rounded and club- shaped, as in Prolecanites. The first pair of saddles have dorsal correspondents and the annular lobe is deep [and acute. The young is costated and the sutures closely resemble those of Phar- ciceras tridens, from which this form is apparently directly de¬ rived. [Triainocerae.] Triainaceras,® nobis, includes but one species which can be separated from Sandbergeoceras only by reason of the transitional condition of the ventral lobe. This has a trident shaped division caused, as in other forms and especially as in Pronorites, by the development of two small pointed saddles on either side of the large funnel lobe, in place of the development of a large median saddle carrying up with it the funnel lobe as in most other genera of Goniatitinae. This genus shows that Pronorites was derived directly from the radical Sandbergeoceras, and that its resem¬ blances to Prolecanites were due to parallelism and. therefore, lecanites, but we find this connection doubtful. Undoubtedly the sutures have quite siiiiilar outlines, but if we compare them with those of Celtites, pi. 28, fig. 5, there is a very close agreement, indicating the same stock rather than the more remote one of Prolecanite;?. 1 H wrinkle. 8 XpCaiva, a trident. 1863.] Í>IT 00 I [Hyntt. could not have been inherited directly from that genus. The only species known to us is the Triai. (Gon.) costatum, D'Arch. et Vern., Geol. Trans. Loud., 2 ser. Vol. 6, pi. 31, fig. 1. This has sinuous outlines in tlie sutures with undivided rounded lobes and saddles, and costated whorls very similar to those of Sandb. tuberculoso-costatum. It can be separated from Pronorites by these same characteristics. Pronorites,^ described by Mojsisovics in Méditer. Trias. Prov. p. 200, includes an exceeding!}^ interesting series of dwarfed forms, which present the marginal divisions of the lobes and saddles sub¬ sequently characteristic of tlie Ammonitinae. Genetic connection with the Prolecanitidae seems to be assured by the asjiect of the sutures. The form of the whorl, and the later larval sutures have the aspect, number of lobes and saddles, and apparently the same mode of developing the outer first pair of saddles from the first pair of saddles, as in the Prolecanitidae. The lobes nvo hastate, the saddles linguiform, the ventral is not fully divided by a si- phonal saddle. The divisions are, in their incipient stages, like minute points or saddles on either side of the large funnel lobes. Thus the apex of the ventral is trilobate. The saddles are rounded, but the first lateral lobes are sul)divided by two incipient sad¬ dles in Pron. mixolobus, according to DeKoninck, a fact not veri¬ fiable in our specimens of this species. There is only one mar¬ ginal saddle in the first lateral lol)e of Pron. cyclolobus. The species, so far as known, are Pron. (Gon.) cyclolobus, sp. Phill. Geol. York. pi. 20, fig. 40-43, mixolobus, ibid, ]>1. 20, fig. 43-47. Popanoceras,^ nobis, includes species of the Dyas, which are very closely allied to Pronorites, but have more com])licated sutures, and approximate more closely to the Ammonitinae. The whorls are more involute and compressed, and are also costated, 1 KoriieSy "Med. Trias. Prov., p. 201, is described by that author as genet¬ ically connected with Pronorites. We are torced to diifer again from this able author¬ ity, since the affinities between these forms are due to larval stuges of the sutures, which are equally characteristic of Gamites, and some other genera. The form of whorl of Norites and the outlines of the sutures appear to us, as to Griesbach, to be closer to those of Sageccras. Norites i^ not very remote from Longobardites, which in our opinion is in th«' young similar to the genus Prolecanite.s both in form oí whorls, and in modes of generating lobes and saddles. It seems to us prossible that the deri¬ vation of the group may have been from tlie lower forms of the Prolecanitidae but not from Pronorites. IXoVavov, a round, flat cake. PROCEEDIKG8 B. S. N. H. VOL. XXII. 22 PEBRUART, 1884. Hyatt ] 338 [April 4^ or marked by furrows. The lobes and saddles are numerous and « club-shaped. The ventral lobes are divided by prominent naiTow, siphonal saddles, carrying small funnel lobes. Three or more pairs of lobes are divided by marginal saddles, either single or double, the terminations of the lobes being either bifid or trifid. Popan, (Gon.) Kingianum, M,, V., K., Russia and Ural, pi. 27, fig. 5, Koninckianuni, ibid, pi. 26, fig. 4, Soboleskyanum, ibid, pi. 26, fig. 5. The extreme form is the Popan. (Arcestes) antiquum, sp. Waagen, Foss. of Salt. Range, Pal. Ind. ser. 13, 1, pl. 1, fig. 10, and this is a close ally of Waagen's Cyclolobus Oldhami.^ [Remarks.] Xenodiscus Waagen, is discoidal and similar to OjDhiceras, Griesbach, Ree, Geol. Siirv. India, Vol. 13, pt. 2, pi. 3, and this is transitional to Otoceras, ibid, pi. 2, which is highly involute. These Dyassic forms are the immediate radicals of the Triassic Ceratitinae. The annular lobes are divided by minute saddles, and the sutures are distinctly ceratitic and cannot be closely com¬ pared with any of the Goniatitinae. The Ceratitinae also in our opinion include, Ilungarites, Cex*atites, Dinarites, Tirolites, Arpad- ites, Beneckia, Meekoceras, Kipsteinia, Balatonites, Trachyceras, Celtites, Badiotites, Proceltites, Lecanites, Gamites, Tropites, Acrochordiceras, Helictites, and Choristoceras. The larvae of these, and many adults, show sutures similar to those of Xannites, Mojsis. pi. 39, and to the Dimerocerae among Goniatitinae. If this view is admitted, the Magnosellaridae and Glyphioceratidae will become the distal Paleozoic radicals of the Ceratitinae of the Dyas and Trias, and the Ammonitinae be confined to Psiloceras and Aegoceras, and their descendants in the Jura. These two genera must be regarded as offshoots of the Prolecanitidae, with either Gyinnites or Monoj)hyirLtes as the immediate radicals. 1 Cycloloba-î, Wangpii, Pal. Ind. ser. 13, 1, p. 21, has for the type Cycl. Oldhami ibid., pl. 1, fig. 9. This genus is very important since it enables us to show the grada¬ tions by which the Prolec.iriitid ie approximate to Arcestes, Ptychites, and Monophyll- ites. Cyclolobus i■•■«•« »v/r; É í. ^ i . '. - » h • # n » t . i' V* #.■ â - •¿*VfA*0'4