250 ON THE rNTEKCENTKUM (JF THE 



ing from the ffont to the back. The iiitercentral articulation is lost, and finally the head 

 of the rib disappears. Only the tubercnlar attachment to the diapophysis remains, 

 in the Streptostyliea, and this is a simple one. 



In the modern Batrachia the rib articulation is that of a double-headed rib to a 

 double-headed diapophysis which originates from the neural arch. There is nothing 

 in the history of the class to show that, as in the case of the reptiles, the head has 

 been transferred or lost. In recent Batrachia, elements resembling the intercentra of 

 Pelycosauria are unknown. The origin of the ribs of fishes and Batrachia have been 

 shown by Goette to be in the connective tissue of the interspaces of the myocommata, 

 which are opposite the middles of the vertebral centra. It follows that there is at no 

 time connection of the ribs with the lines of separation of the vertebral bodies. 



If the intercentra become the functional centra in the Sphenosaurid.T, it is likely 

 that they are such in the Stegocephali, and in the modern Batrachia.*^ It follows, 

 then, that Batrachia, excepting Khachitomi and Embolomeri, have no centra, but in- 

 tercentra only. This view is confirmed by three facts, two of them already mentioned: 



I. There are no intercentrum-like bodies in existing Batrachia. 



II. The ribs which originate from intercentral cartilage and intercentra in reptiles, 

 originate from the principal vertebral bodies in Batrachia. 



III. The chevron bones are continiia with intercentra in Reptilia, and with the 

 caudal vertebral bodies of Batrachia. 



With regai'd to the last proposition I may add, that I have examined young Nec- 

 tuncs maculatus, of four inches in length ; of larvfe of Gurinopliilus j^orphi/viticus, of 

 thi-ee inches ; of Spelerpes ruher of two and a half inches ; of Amhlystoma yuactatum, 

 of 35mm ; and of Siyderpes hilmeatus, of 25mint length. In none of these is the 

 least trace of articulation of either chevron bone or of neural aix-h in any part of the 

 column to be discovered (Plate I, Fig. 11). 



Researches into the embryology of the Urodela and Anura have not yet brought 

 to light any traces of the rhachitomous structure ; which is probably a case of inexact 

 parallelism, cceuogeny or falsification of the embryonic record — a phenomenon which 

 is not uncommon. There can be no doubt, however, that the entire record was i)re- 

 sented in the embi-yonic history of Permian land Vertebi-ata. and for a long period 

 subscquentl}', but that the rhachitomous stage has been, with the true centrum, lost 

 from tiie batrachian line at least. 



As I have shown, in the Embolomeri (he intercentra tend towards reduction in 

 the dorsal i-egion, while the centra arc predominant. If the tendency of the evolution 



* American NiitunUisI, 1880, p. 77. 



t For these specimens I am greatly indebted to the Smithsonian Institution and the Secretary, Prof. Bairil. 



