Notes on the morphology of the Tunicata. 551 



while in my specimens they lie above. I can only suggests that 

 Herdman has failed to find the ciliated funnel, which in my specimens 

 lies under the front part of the brain and well down in the con- 

 voluted walls of the "oesophagus", and that he has mistaken some 

 other structure, possibly one of the many folds of the oesophageal 

 wall, for the funnel. He does not show the rapheal nerve, unless, 

 possibly, it be that which he figures as the duct of the gland. The 

 well known accuracy of Herdman's observations upon the Tunicata 

 makes me hesitate greatly to suggest that he is mistaken in his de- 

 scription of these organs, yet it seems hardly possible that two species 

 so evidently belonging to the same family and probably to the same 

 genus should differ so greatly as Herdman's descriptions and my own 

 would indicate. I know of no other description, besides this of Herd- 

 man's for Ocfactnemus hifhyus, which shows, in any species, the funnel 

 lying behind the ganglion and sending its duct forward over the 

 gland toward the ganglion. Such a condition of affairs would indicate 

 a most profound distortion of the usual space relations. 



It needs no discussion to show that the neural gland of Oct- 

 acnemus conforms to the type found in the Clavelinidae and the com- 

 pound Ascidians (exclusive of the BotrylUdae), rather than to the 

 type seen in the Salpidae. In Section V of this paper, I discuss the 

 anatomy of Octacnemus, showing reasons for believing that it is to 

 be regarded as a social Ascidian, rather than as a form related to 

 the Salpidae. 



The Apx>endiculariidae, 

 Plate 40, Fig. 80. 



In most members of this family no neural gland has been de- 

 scribed. In my own specimens of an undetermined species I have 

 found the conditions much the same as those described by Fol^) for 

 FritiUaria. No neural gland is present but a trumpet-shaped, ciliated 

 funnel lies at the right side of the anterior end of the nerve tube and 

 sends its upper end up along the right side of the sensory vesicle to 

 end blindly at a level considerably above the dorsal surface of this vesicle. 



Chun '^) found in Megalocercus ahyssorum that the tip of the 

 funnel is prolonged backward into a tube lying dorsal to the nerve 

 tube (Fig. 80). Upon this little prolongation of the funnel were two 

 swellings. The nature of this organ was not determined by study of 



1) Fol, 1872. 



2) Chun, 1881. 



