358 JOHNS HOPKINS UNIVERSITY MORPHOLOGICAL MONOGRAPHS. 



ences are the persistence of the neural canal in the adult pyrosoma 1 and 

 the very slight degree of thickening of its ventral wall. The latter of 

 these is a point of difference from all other tunicates and is probably to 

 be regarded as secondary. In comparing the nervous systems of the 

 ascidians, pyrosoma, doliolum and salpa, we find the latter three much 

 more alike than either of them is to the ascidians. Of the latter three, 

 salpa appears to present the most resemblance to the ascidians, because 

 of the presence in these two alone of a subneural gland with paired 

 lateral ducts; but this point is not one from which alone relative degrees 

 of relationship can safely be deduced, for the same structures may once 

 have been present in pyrosoma and doliolum, though now they are lost 

 in these genera. The ease with which they may be lost is shown by the 

 fact that in the chain Salpa democratica-mucronata they have disap- 

 peared. The presence of these structures in ascidians and salpa does 

 strongly indicate that the common ancestor of the two had a subneural 

 gland with one or more pairs of lateral ducts. Emphasis is given to this 

 view by the fact that in Salpa africana-maxima, chain form, whose 

 nervous system shows the most primitive character, we have the greatest 

 development of the gland. 



The lateral ducts of the subneural gland in Phallusia mammillata 

 connect with the peribranchial chamber, according to Van Beneden and 

 Julin, while in salpa the lateral ducts connect with the pharynx or 

 branchial chamber. This difference at first sight seems to preclude the 

 homology of these organs in the two cases. But still, after careful study, 

 it seems to me the structures may be truly homologous for several 

 reasons. They have a decided superficial resemblance. They are located 

 in the same region with reference to the ganglion. They have the same 

 function ; at least in both cases they are gland ducts. It is furthermore 

 easy to conceive a shifting of position by which the lateral ducts, origin- 

 ally opening to the peribranchial chamber, came to open to the branchial 

 chamber. The peribranchial and branchial chambers in salpa are later- 



1 Salensky [14] regards the canal opening into the funnel of the adult pyrosoma 

 as a new structure and not as a remnant of the canal of the central nervous system. 

 Whether it be a new structure or the original canal reopened, there is at any rate a 

 closing and at least partial degeneration of the canal of the central nervous system 

 previous to the appearance of the definitive funnel. This degeneration is very 

 probably homologous to the similar degeneration of the anterior portion of the canal 

 of the central nervous system observed in salpa and doliolum (and probably also to 

 the degeneration of the duct leading from the gland to the funnel seen in some indi- 

 viduals of the ascidian, Phallusia mammillata). 



