186 University of California Publications. [zoology. 



possession of sufficient facts to justify a positive opinion as to 

 the value of the suggestions made at that time. We prefer, 

 consequently, to do no more now than make the following 

 remarks: First, the esophageal baud is not, as might be 

 inferred, structurally wholly unique to this tornaria. Morgan, 

 '94, has since figured, PI. Ill, Fig. 5 of this paper, what is 

 unquestionably the same thing in the esophagus of the Bahama 

 tornaria, and in all likelihood the ventral wall of the esophagus 

 is somewhat thicker than the lateral walls at least, in all species; 

 and it is further probable that the esophageal cilia will be found 

 to be confined to the ventral and dorsal sides with a preponder- 

 ant development on the ventral side in most if not all species. 

 Second, the doubt expressed in the preliminary paper as to 

 whether the ventral esophageal band is a direct continuation of 

 the ventral gastral band may now be removed. The two can be 

 looked upon only as parts of one and the same band, though 

 with different degrees of development in the two parts, the 

 stomach part bearing stronger cilia, and being more strongly 

 differentiated from the enteric wall in general, than the esopha- 

 peal part. Third, the statement that the esophageal band 

 probably has the same function of the endostyle in the Tunicates 

 should be modified so as to include the esophageal and gastric 

 portions together instead of being restricted to the esophageal 

 part alone. Thus modified we would reassert this view, and 

 point out in addition that the particularly large cilia of the gas- 

 tric band and the posterior end of the esophageal band suggest 

 the well known remarkably long cilia of the floor of the tunicate 

 endostyle; and furthermore that in the gastric portion of the 

 band, at least, there are almost certainly some secretory cells. 



Fourth, and finally, while our renewed study of this point, so 

 far as we have been able to carry it, has strengthened the sug- 

 gestion of a general functional similarity between this band 

 and the proehordate endostyle, it leaves the question of true 

 homology as doubtful as ever. 



((') ECOLOGY. 



We may now turn to the habits and activities of the tornaria. 

 During at least the first and second larval periods the life 



