der of the Ordovician of this country was, as we shall see, taken up and passed on by the laticosta and costata types.

The fact that a more nearly normal form of lynx survived to the close of the Ordovician (Madison beds) has already been pointed out. This form seems to represent the final expression of the type-a shell, primitive in many respects, though of large size; and undergoing a last attempt to accommodate itself to changing conditions.

Platystrophia laticosta.*—This is one of the most interesting and one of the least understood types of Platystrophia. It seems to be confined to American faunas, and occurs here only in the Cincinnati group of the Ohio valley, where it ranges through the Lorraine and reappears at the base of the Richmond formation.

Meekt has admirably described this variety and little can be

FIG. 16. A and C, Platystrophia laticosta from Vevay, Indiana; ×%. B, P. costata from Cincinnati, Ohio; . A, shows growth stages I-V separately drawn (b-f), and anterior (g) and posterior (h) views: note the disappearance of the right hand plication of the sinus after stage V. B, shows the initiation of a second and third plication in the sinus at stage II (a, b) and their disappearance at stage III (a, c). C, is a gerontic individual of P. laticosta; note the increase of the cardinal angle after stage I, and the reduction in the height of the fold, and strong growth varices (c, d). A and C, author's collection; B, Yale collection.

added in regard to its adult characters. Its minute study, based upon hundreds of individuals, has, however, developed unsuspected relationships, which I shall now point out. Fig. 15 gives

* James, Cat. L. Sil. Foss. Cincinnati Group, 1871, p. 10.

+ Pal. Ohio, i, 1873, p. 116, pl. 10, fig. 4. P. chama de Verneuil (not P. costata Pander) may be related to P. laticosta.

some idea of the variation of laticosta; e is a normal individual with one large (median) and two small (lateral) plications in the sinus. a has but one plication in the sinus and the fold is extremely elevated. b to d are intermediate between these two. In seeking the derivation of this type as in other types of Platystrophia the growth stages are of paramount importance. In fig. 15, c, are shown two lateral plications in the sinus originating at the usual distance from the beak but disappearing a little over half way from the beak to the front margin. Fig. 16, A, shows the inception of this process of reduction of the lateral plications of the sinus. Here only one plication is affected, and that only near the front margin. Fig. 16, B, shows the same tendency in P. costata.

Considering now the growth varices (fig. 16) it is evident that laticosta becomes progressively more transverse during ephebic stages, while costata becomes progressively less transverse; the early stages of the two being identical and alsoidentical with the early stages of P. lynx. In P. laticosta the cardinal angle is progressively 72°, 73°, 63°, 68°, and 70°, returning thus in gerontic stages to near the angle of an early ephebic stage. In costata (fig. 16. B) the angle changes from 95° to 78° to 82° to 99° in the fully adult stage, while in gerontic stages it may be as much as 113°. The largest angle seen in any stage of laticosta was 95°. Both laticosta and costata are therefore derived from a primitive lynx by a reduction of the number of plications, and an elevation of the fold. The relation of laticosta to costata is not a linear one; but after the establishment of a pauciplicate stock like the young of both (see ante), a divergence occurred, one branch taking the direction of an elongate narrow type (costata) and the other the direction of a transverse acuminate type with extremely high fold (laticosta).

The above mentioned uniplicate laticosta (fig. 15, a) is found in the Upper Lorraine and is not to be confused with P. costata. The former has an exceptionally high fold but the same contour as a normal laticosta, from which variety it is derived, and with which it is connected by every possible gradation. If it is desired to distinguish this form from P. laticosta, it may conveniently receive the name of unicostata.* In the Dalmanella Meeki zone which intervenes between the Lorraine and Richmond beds, no specimens of Platystrophia occur, except at the very top of the zone where the typical

*The dentata and crassa of American authors undoubtedly include these two forms costata and unicostata. The former term will be shown to apply only to certain foreign and Upper Silurian biplicate types. Since the term crassa confuses two distinct forms it had better be abandoned altogether, especially since the form to which it was intended to apply is the costata of Pander.

FIG. 17. Group of Flatystrophia acutilirata from the base of the acutilirata zone (Lower Richmond) Tanner's creek, Indiana, with forms transitional to P. laticosta from the same zone. aa', P. laticosta, after Meek; bb', P. laticosta, Tanner's creek, lower Richmond; cc', intermediate acutilirata type; dd', very mucronate acutilirata; ee', normal acutilirata. Author's collection.

Within the next 20 ft.

laticosta again makes its appearance. of strata this laticosta is modified into a typical acutilirata. Fig. 17 will make this clear. The transition has been noted in so many individuals, from so many different localities, that there can be no doubt as to the correctness of this view of the relationship of these two forms. If any additional evidence were needed, it is furnished by a study of the early stages of acutilirata and by the general angularity and high fold of the latter species; as well as by the absence of the laticosta type from the Rhynchotrema zone, where acutilirata abounds.

Few individuals of laticosta present pronounced gerontic modifications. Such changes when they do occur produce a shell of extreme gibbosity, and with a large cardinal angle (from 65° to 74° in one specimen), so that the contour of the shell approaches that of P. lyna. Fig. 16, C, illustrates this. Normal growth ceased at stage I: the fold subsequently becomes lower relative to the size of the shell (cf. stages I and II), and the frontal profile becomes regularly curved instead of being truncated as in ephebic stages (cf. fig. 16, Cc, with fig. 15, e"); so that the profile also resembles that of a gerontic

P.lyne. In fact the gerontic laticosta, in almost every feature in which it departs from the normal adult type of the variety, approaches P. lynx.

In another place* I have pointed out the intimate connection between Platystrophia laticosta and P. lynx. The examination of many thousands of specimens of these forms has failed to bring to light any character which does not show transitional stages from one to the other. The relatively greater strength and smaller number of the plications of laticosta is its most constant character, and with this is usually combined auriculation and prolongation of the cardinal extremities the latter character is, however, by no means uncommon in P. lyne. It seems best therefore, to the writer, to consider laticosta as a variety of lynx, transitional between it and the species acutilirata.

For comparison with the uniplicate laticosta I have inserted (fig. 15, h) figures of P. chama traced from de Verneuil's figures. That author says (op. cit.) he succeeded in obtaining a complete series of transitional forms between the small narrow type of chama (= costata Pander) and the acuminate type figured. As I have not seen any specimens of P. chama, de Verneuil's statement must be taken as indicating the relation of these two forms; though I strongly suspect that chama may bear no closer relation to costata Pander than does the uniplicate laticosta. The latter may, indeed, be the form to which it more nearly corresponds.

Platystrophia acutilirata.+-Another type, found only in American faunas, is Platystrophia acutilirata Conrad, which is confined to the Richmond beds of the Cincinnati group.§ This species presents the most remarkable similarity to a Spirifer of any of the types of Platystrophia. The breadth may become as much as three times the length; and the cardinal extremities are frequently as acuminate as those of an average Spirifer mucronatus, yet between these extreme forms * This Journal, July, 1902, p. 14.

+ Géol. de la Russie, 1845, pl. v. figs.1b, 1c.

Conrad, Jour. Acad. Nat. Sci., Philadelphia, viii, 1842, p. 260.

Keyes lists this form from the Hudson shales (Cincinnati) of Louisiana, Missouri. As he gives no figure or description, it is uncertain whether his specimens are really referable to acutilirata or belong to the acuminate type of lync. See Keyes, Geol. Missouri, vol. v, Paleontology, 1894, p. 66.

Prof. H. S. Williams has called my attention to the fact that Atwater's specimen of S. pennatus came from a locality in Ohio to which examples of Platystrophia acutilirata might have been transported by streams. While I do not believe that Spirifer pennatus is a Platystrophia acutilirata, the possibility of this being the case together with the total inadequacy of Atwater's description and figures, would seem to warrant abandoning his name altogether and returning to the well known name of mucronatus for this species of Spirifer. See Atwater, this Journal, ii, 1820, p. 244, pl. I, figs. 2, 3.

AM. JOUR. SCI.-FOURTH SERIES, VOL. XV, No. 85.-JANUARY, 1903.