M. M. METCALF ON THE EYES AND SUBNEURAL GLAND OF SALPA. 349 



ing to Salensky, the primitive ciliated funnel and the duct connecting it 

 with the brain atrophy, and later the definitive ciliated funnel arises at 

 the spot where the primitive funnel at first appeared. The posterior end 

 of the definitive funnel elongates till it comes to lie along the ventral 

 surface of the ganglion, forming in the adult a hollow tube opening 

 anteriorly to the ciliated funnel. It seems probable that the definitive 

 funnel is merely the primitive funnel reopened, but be this as it may, 

 the primitive funnel must probably be regarded as the homologue of 

 the funnel in salpa and doliolum. This is shown by the perfect agree- 

 ment in the three genera of the condition of these organs in the imma- 

 ture individual, and also by the greater or less degeneration of the com- 

 munication between brain and funnel which occurs in all three genera 

 subsequent to the time when they so closely agree in the structure of 

 these organs. (See Appendix I.) 



In the ascidians the early relation of the ciliated funnel and the 

 duct that opens into it to the nervous system is difficult to determine, 

 because of the contradictory statements of those who have studied these 

 organs in the group. Kowalevsky (9) and others say the canal of the 

 central nervous system at an early stage opens to the pharynx. From 

 analogy with pyrosoma, doliolum and salpa this would be what we 

 would expect. Van Beneden and Julin (1), however, deny the presence 

 of such an opening, saying that the posterior prolongation of the ciliated 

 funnel does not communicate with the canal of the central nervous 

 system. According to these investigators the canal of the central 

 nervous system in the larval ascidians is wholly distinct from the 

 ciliated funnel. The nervous system is divided into three regions: (1) 

 an anterior sense vesicle which atrophies in the later development; 

 (2) a visceral portion from whose ventral cells the gland develops, while 

 its dorsal cells give rise later to the adult ganglion ; (3) a caudal portion 

 which degenerates completely. In the later larval development, the 

 posterior prolongation of the funnel, which is short in the earlier stages, 

 elongates considerably, pushing between the ganglion and gland and 

 becomes the efferent duct of the gland. 



It is difficult to form an opinion whether in the early ancestors of 

 the tunicates the canal of the then simple nervous system opened to 

 the pharynx, and this connection was lost in the larval ascidian, owing 

 perhaps to the great development of the anterior sense vesicle; or 

 whether the ciliated funnel arose separately from the nervous system, 

 as some claim it now does in the ascidians, and a communication 



