i894- 



SOME NEW BOOKS. 



461 



rise to a little two-walled sac or gastrula. In many groups of verte- 

 brates this simple form of development becomes impossible ; for the 

 egg-cell becomes so loaded with inert food-yolk that it does not divide 

 completely, but the shape of the embryo is modified precisely in pro- 

 portion to the amount of food-yolk present. The yolk is not 

 distributed in the same way in all the animals possessing it. In some 



Fig. 6. 



Fig. 7. 



Fig, 6. — Egg of Distaplia : the first division has occurred, and follicle-cells are 

 growing in between the segmentation spheres. 



Fig. 7. — Diagram of a blastosphere where follicle-cells are passing into the 

 segmentation cavity. 



cases it is restricted chiefly to the cells vv^hich are to form the 

 alimentary canal. In others, as anyone who has examined a young 

 tadpole must know, all the cells are packed with yolk-spheres. In 

 Pyrosoma, a near ally of Salpa, there is a large amount of food-yolk, 

 and that is massed not in the future digestive-cavity but in the body- 

 cavity. But the food-yolk is aided by follicle-cells, which push their 

 way into the developing cells and are absorbed by them. Dr. Brooks 

 suggests a method by which the peculiar arrangement in Salpa might 



Fig. 8. Fig. 9. 



Fig. 8, — Diagram of invagination in an embryo such as that of Fig. 7. 

 Fig. 9. — Diagram of the gastrula of an embryo in which the follicle-cells form 

 an epithelium. 



have come about. Fig. 6 represents an actual stage in the develop- 

 ment of Disiapha, which has cells bloated with food-yolk, but still able 

 to divide completely. Two segmentation spheres are represented, sur- 

 rounded by a number of follicle-cells, some of which are pushing their 

 way in between the segmentation spheres, and which are absorbed as 



