342 BULLETIN OP THE 



than in the adult Salpa, and shows that in this respect the embryo 

 resembles the adults of other tunicates.) As soon as this locomotive 

 life had been established it would be of advantage to have as much 

 water as possible pass through those branchial slits which lie in the 

 straight line connecting the two external apertures, aud accordingly 

 those at the base of the branchial sac would become excessively devel- 

 oped at the expense of those upon the sides, and these latter, being 

 no longer of functional importance, would tend to disappear, and if 

 the w r alls of the lateral atria were at all muscular, as they undoubt- 

 edly are in all Tunicata, we can easily understand how they might be 

 so modified as to violently expel the water from the body, and thus 

 still further assist in locomotion. If Salpa is descended from a form 

 having the ordinary branchial sac the presence of the lateral atria in 

 the embryo can be understood, but if we deny all evolution or consider 

 Salpa a low form, they are as meaningless as the aortic arches of an 

 embryonic mammal would be if we did not refer them to a gill-bearing 

 ancestor. 



In one respect Salpa is more embryonic than the fixed Ascidians. 

 The structure of its ganglion and sense organs is much more special- 

 ized than in them, and resembles that of their larvae in this respect ; 

 but we can see that while a fixed animal, having little need of a 

 specialized nervous system would be likely to possess oue much more 

 rudimentary than that of its locomotive larva, a form which re- 

 mained free throughout life would be likely to retain it in its highly 

 developed form. 



We come now to the question : Will the theory that Salpa has been 

 adapted to a locomotive life throw any light upon the separation of 

 the sexes 1 



" It must have struck most naturalists as a strange anomaly, that 

 both with animals and plants, some species of the same family and 

 even of the same genus, though agreeing closely with each other in 

 their whole organization are hermaphrodites, and some unisexual." 

 In Salpa we have such a case, which a comparison with the other 

 Tunicata shows to have been originally composed of two hermaphro- 

 dites ; the male organ of the solitary form having been converted 

 into the elaeoblast, which becomes excessively developed, and sup- 

 plies the material for the formation of the chain, while in the chain- 

 salpa the ovary has entirely disappeared. 



