Development of (he Mammalian Phylum. 367 



to them in its earliest representatives, a second order 

 appeared, that of the Theroniorpha, which exhibits an extra- 

 ordinarily many-sided development, and to wliich we have 

 to devote closer attention. Likewise referable to the Progono- 

 sauria are the two orders Sauropterygia and Ichthyosauria, 

 which enjoyed a pelagic existence upon the open sea and 

 have undergone profound transformations in their structure, 

 quite analogous to those which at a later epoch of the earth's 

 history were experienced by the whales among the mammals. 

 Very old also is the order of the Crocodilia, a branch of which 

 has been preserved until the present time. In consequence 

 of the palteontological facts their pliylogeny is considered to 

 be very well understood. The earliest crocodiles are Triassic ; 

 then forms greatly changed in aspect reappear in the upper- 

 most Jurassic formation ; these are traceable through all 

 subsequent strata up to the present time. Now it is highly 

 instructive to see how incomplete are even the best pala3on- 

 tological records, for, on the basis of embryological investi- 

 gations upon crocodiles, I am driven to conclude that their 

 ancestors were at a certain time pelagic animals * with corre- 

 sponding characteristic morphological peculiarities, and only 

 gradually developed into the littoral and fluviatile creatures such 

 as we find them to-day. Pala3ontology, however, has no know- 

 ledge of such pelagic ancestors ; its attention is first directed 

 thereto by means of embryology, and it is to be hoped that 

 we shall one day succeed in finding remains of the supposed 

 ancestors in the strata which precede the uppermost Jurassic 

 series. With the oldest crocodiles as well as the Progono- 

 sauria (Rhynchocephala) there is connected an order which 

 excites general interest owing to the fact that it contains the 

 largest terrestrial forms that the earth has ever {)roduced. 

 The length of the American Atlantosaurus^ for instance, 

 amounted to 115 feet, and its height to 30 feet, while its 

 thigh was more than G feet long, and at its ui)per end exceeded 

 2 feet in diameter. Since these animals employed exclu- 

 sively the ])ind legs for walking, a transformation of the 

 hinder extremities, as also of the jiclvis, was produced owing 

 to the transference of the weight of the body to them, just as 

 we see is the case in the birds in consequence of the same 

 physiological cause. In spite of the fact that we may not 

 at once utilize such resemblances for the purpose of estab- 

 lishing a phylogenetie connexion between the two, it is never- 

 theless conceivable that Dinosauria and birds may have 



• lu a paper wliieli is at present in tho press I have endeavoured to 

 prove this assertion by means of the embryology of the skeleton of the 

 hand. 



