November 27, 1896.] 



SCIENCE. 



771 



both the observed types there is practically 

 no cardinal plate, the hingeline being thin, 

 the ligament between (not outside of) its 

 opposed edges and usually directed ob- 

 liquely downward and backward. Toward 

 the extremities of the cardinal border are 

 found the feeble projections which indicate 

 the beginnings, called by Bernard the primi- 

 tive lamellce of the permanent teeth, of which 

 the anterior appear first. In the left valve 

 one, and in the right valve two, of these 

 appear, from which are developed ultimately 

 the anterior lateral and the cardinal teeth. 

 The posterior lamellae give rise to the pos- 

 terior laterals only. 



The position of the ligament above de- 

 scribed, and which seems to be general at 

 this stage, is regarded by Bernard and at 

 first sight would seem antagonistic to the 

 dynamic hypothesis ofNeumayer and others, 

 which postulate an originally external liga- 

 ment. A little reflection, however, shows 

 that there is no real antagonism, for there 

 is practically no mechanical distinction be- 

 tween inside and outside at this stage, and 

 in the later stages it makes no essential 

 difference, so far as the dynamics of the 

 hinge are concerned, whether the move- 

 ment which leads to a separation between 

 ligament and resilium is upward and out- 

 ward for the former, or downward and in- 

 ward for the latter, the mechanics of the 

 process being the same in either case. The 

 ligament increases by additions from below, 

 or at the posterior end. If these are in ex- 

 cess in the latter case the ligament tends to 

 become elongated and external, in the for- 

 mer case short and internal ; while its edges 

 of insertion, through the deposit of shell ad- 

 jacent to them, in a ligament otherwise ex- 

 ternal are almost always situated in a 

 channel of which the nymphae are the 

 thickened ventral border. The dynamic 

 reaction of the form and mode of growth of 

 the ligament upon the form of the valves is 

 clearly set forth by Bernard, who thus sup- 



plies an interesting contribution to the data 

 of dynamical evolution. 



The most important and unexpected re- 

 sult of Bernard's studies is the discovery 

 that, in the nepionic stage oi all the Prionodes- 

 macea, the first development of the hinge 

 consists in the appearance, on each side of 

 the ligament (except in Ostrea, where they 

 are solely posterior) , of a series of vertical 

 crenulations, or taxodont denticles, which 

 are entirely distinct from the permanent 

 teeth of the adult shell. These had been 

 observed in a few cases previously, but 

 their widespread occurrence, not only in 

 the order mentioned, but also in some 

 species of the Teleodesviacea, had not been 

 suspected ; while the discontinuity between 

 them and the permanent teeth, even in 

 such typically taxodont groups as Nucula 

 and Pectuncidus, is very remarkable. For 

 this primitive hinge, as distinguished from 

 the subsequently developed permanent car- 

 dinal mechanism, we may adopt the term 

 provincidum. The character of the provincu- 

 lum is that of two subsymmetrical areas 

 of nearly vertical parallel ridges separated 

 hy subequal grooves, forming a taxodont ap- 

 paratus similar to the permanent dental 

 series of Area and much resembling the 

 hinge in some of the Paleoconcha illus- 

 trated by Neumayr. It is difl&cult to avoid 

 the conclusion that we have in the provin- 

 culum a representative of the primitive 

 hinge of the Protopelecypod, which was, 

 perhaps, a free-swimming pelagic animal 

 like Planhtomya. The theory of N'eumayr, 

 which derives the dentition from the influ- 

 ence of external sculpture on the hinge 

 margin, accentuated by natural selection, 

 remains unshaken, but we have still to ac- 

 count for the gap which, in the forms yet 

 examined, appears to intervene between the 

 provinculum and the permanent teeth. It 

 is probable that further researches will lead 

 to the discovery of this missing link. 



The second important fact brought out 



