THE PALEONTOLOGIC RECORD 151 



and character in the ancestral type at the same stage of its development, 

 it represents an actual repetition of past history and is therefore palin- 

 genetic. Sometimes it is not quite clear, however, under which caption 

 the embryonic structure comes, and its interpretation must be attempted 

 with caution. 



Osborn in his lectures to his students speaks of the three-fold evi- 

 dence for evolution which stands firmly like a tripod, the legs of which 

 are comparative anatomy, embryology and paleontology; and the evi- 

 dence of each should correspond, provided the interpretation be correct. 

 Of these, however, embryology is manifestly the weakest member, while 

 paleontology is a tower of strength ! 



The reptiles are so rare as embryos and withal so ancient a group 

 that their ontogeny throws but little light upon paleontology. Among 

 the fossil forms a number of specimens of Ichthyosaurus have been 

 found with young contained within the body of the adult. Many of 

 these are in the normal position for fceti-in-utero ; others are displaced, 

 with the head directed forward. These latter Branca thinks may be 

 young that have been eaten. There is also, at times, a very great 

 difference in the size of the contained young. Aside from a slight dif- 

 ference in proportions, especially that of head to trunk, and a less degree 

 of hardness of the embryonic bones, as indicated by their being crushed 

 over the parent's ribs, the young teach us nothing as to ancestral struc- 

 ture as they are in every way perfect ichthyosaurs. They do prove, 

 however, when the evidence of viviparity which they offer is taken in 

 connection with the supreme degree of aquatic adaptation indicated, 

 that the ichthyosaurs were high sea-forms, never coming ashore even 

 for egg-laying. 



That certain of the dinosaurs were also viviparous may be proved 

 by an embryo contained in the unique specimen of Compsognathus 

 longipes from the Jurassic of Bavaria. So far as I am aware this 

 embryo gives no other evidence of ontogenetic value. 



The turtles have been made the subject of exhaustive study by Hay 

 and from the embryological point of view by Clark under L. Agassiz. 

 Anatomically they are the most remarkable of reptiles, having under- 

 gone during their career an extreme modification in many directions 

 while retaining a number of very primitive characters. The most 

 remarkable feature is the development of the shield or carapace, which 

 contains what are generally considered as the homologues of the ribs of 

 other vertebrates, but, strangely enough, here lying outside of the 

 shoulder girdle, a feature wherein the turtles are utterly unique. The 

 embryology, which is well known, ought to throw some light upon the 

 origin of this important feature. In their earlier stages of develop- 

 ment the Chelonia resemble the Lacertilia, the chief pecularity being 

 caused by the development of this carapace which appears in the form 



