M. M. METCALF ON THE EYES AND SUBNEURAL GLAND OF SALPA. 357 



If then we homologize the dorsal part of the ganglion in salpa with the 

 adult ascidian ganglion, and the ventral third or quarter of the salpa gan- 

 glion (the original thickened portion of the ventral wall of the neural 

 canal) with the thick ventral wall of the visceral portion of the larval 

 ascidian nervous system, we are justified in saying tliat the ascidian gland, 

 which is formed from the thick ventral wall of the visceral portion of 

 the larval nervous system, is represented in salpa by the ventral portion 

 of the ganglion. Of course this never functions in salpa as a gland. The 

 homology is strengthened by the presence of the lateral tubes in salpa, 

 which connect the two hollow disks upon the ventral surface of the 

 ganglion with the branchial chamber. These seem to be very closely 

 related to the communications between the subneural gland and the 

 peribranchial chamber, found in Phallusia mammillata. The com- 

 parison is made more perfect by Herdman's description [4] of two indi- 

 viduals of Phallusia mammillata in which the duct of the subneural 

 gland leading to the ciliated funnel had atrophied. 



The bearing of this Jwmology upon the relation of the eye of salpa to the 

 larval ascidian eye and the pineal eye of vertebrates is evident at a glance. 

 Salpa eye is a new structure wholly unrelated to the ascidian eye. It 

 develops from cells which have their origin in an entirely different 

 region of the nervous system. It cannot be homologized with either of the 

 epiphyses of the vertebrate brain, for it does not arise directly from 

 the central nervous system as do the latter, but arises from a secondarily 

 acquired, modified portion of the nervous system, not represented in the 

 vertebrate brain. For the same reason the pigment spots in the ventral 

 part of the ganglion of pyrosoma cannot be homologized with the larval 

 ascidian eye or the pineal eye of vertebrates, as Seeliger [17, p. 582] 

 claims, but they are more probably related to the posterior or postero- 

 ventral eyes of such species as Cyclosalpa pinnata or Salpa costata- 

 Tillesii. 



Does the condition of the nervous system and subneural gland of 

 salpa furnish any evidence as to the relationship of salpa to the other 

 tunicates ? I have shown the almost complete resemblance between the 

 nervous systems of salpa and doliolum at a certain stage of their develop- 

 ment. The, nervous system of salpa passes through a doliolum stage. This 

 would be evidence then, so far as it goes, that doliolum and salpa had a 

 common ancestor and that doliolum stands the nearer to that ancestor. 

 The nervous system of pyrosoma, though a little more removed from that 

 of salpa, still very closely resembles it. The two most important differ- 



