128 COMPARATIVE ANATOMY. 



long, tailed embryo* is developed, which escaping from egg 

 moves freely in the water. After a few hours it becomes 

 fixed, loses its caudal appendage, and without further change 

 is developed into the sexual form. 



The phenomena of alternate generation are witnessed in 

 Salpa. The egg undergoing complete segmentation evolves 

 a larva which assumes a form recalling that of the parent, to 

 which it remains for some time attached by a placenta-like 

 structure. Rupturing the connection, a curious process of 

 gemmation, which was commenced while the larva was yet 

 confined, is completed. This consists in the formation from 

 a central cylindroid stem (stolon) of numbers of buds, all of 

 which being evolved at about the same time * adhere together 

 by their external surfaces or special organs of attachment' 

 forming chains or clusters of sexual individuals. These are 

 generally composed of two rows joined together by several 

 cords and enveloped in a common membranous tube. The 

 sexual form is known as the < aggregated Salpce' or 'stock,' 

 the larval or non-sexual, as the ' solitary Salpa.' 



In Pyrosoma, the yolk is diffused within ovisac at an early 

 stage of development, and while the embryonal mass is yet 

 newly formed, it partially segments transversely into five 

 divisions. The first of these is a provisional structure; the 

 other four are permanent ones, and each subsequently at- 

 tains the sexual form. While within the parent they are 

 joined together by a connecting band. The subsequent 

 growth of each of the embryos into a compound animal has 

 not been traced. 



Brachiopoda. Development but little known. In Crania 

 the young possesses four pairs of ciliated processes, situated 

 around the mouth, and which, together with the gullet, can 

 be protruded from the cavity of the shell. It bears a strong 

 resemblance to the permanent condition of Polyzoa. 



Lamellibranchiata. Egg not uncommonly possesses a nii- 



* To avoid multiplication of terms 'embryonal' and 'larval forms' will 

 be used indifferently, though in so doing a lack of precision is acknowl- 

 edged. 



