36 J. GRAHAM KERR. 



egg of Rana palustvis), or Dean's^ pi. iv, fig. 62 (Aci- 

 penser, longitudinal section of egg dui-ing gastrulation), 

 to see the remarkable unity which runs through these dif- 

 ferent types. Of the figures which I happen to have men- 

 tioned, the first three might have been used to illustrate the 

 corresponding stages of Lepidosireu or Protopterus 

 almost as well as the figures which I have given. 



Looking at the broad facts in these three groups, and com- 

 paring them with what occur in other vertebrates, one cannot 

 but be struck with the fact that in the only two groups in 

 which it is almost certain that we have to do with poorly- 

 yolked eggs in forms descended from richly-yolked ones, viz. 

 the Teleostei and the higher Mammalia, we find that in neither 

 has the process of gastrulation reverted to its original cha- 

 racter. Rather by its profound modification from the normal 

 type it bears witness to the changes which have taken place 

 in its history. This being so, the comparatively simple type 

 of gastrulation similar in Petromyzonts, Amphibia, Dipnoi, 

 and Ganoids cannot but weigh strongly as evidence against 

 the view propounded by Rabl, that any of these groups 

 are descended from ancestors with large, richly-yolked 

 nieroblastic eggs. 



There is one point which I should like, in conclusion, to 

 draw attention to, and that is the shunting forwards in 

 development of the rudiments of organ systems to an earlier 

 period than that to which they normally belong. Thus by 

 Stage 14, when gastrulation is just completed, the study of 

 sections teaches us that the embryo is already a compli- 

 cated triploblastic organism, with definite mesoblast and 

 chorda. 



I have obtained the small results recorded above only by 

 prolonged work upon a most extensive material preserved 

 with the greatest care and by the most approved methods. 

 I have been greatly impressed by the variability observed 

 amongst embryos of similar stages in development, much of 

 it probably natural, much of it certainly due to differences 

 1 *J. Morpliol.,' vol. xi. 



