728 CROZIER. 



ing them upon A. atra in small aquaria containing no other tunicates. 

 But they would not remain on this ascidian, they would not feed, and 

 indeed died in the course of two days. 



There consequently seems to be a close correlation between the 

 habits of the three color varieties for I believe the evidence warrants 

 the view that they are "physiological varieties" of one species and 

 their ability, on the one hand, to feed on the particular ascidian which 

 harbors them and, on the other, to "make use" of the pigments 

 involved. This correlation is also suggested by the size of the three 

 varieties, which, so far as I have been able to observe, increases 

 directly with that of their tunicate hosts (E. turbinata, A. curvata, 

 A. atra), the average lengths of the respectively associated polyclads 

 being 6, 17, and 20 mm. 



AGAR'S ISLAND, BERMUDA. 



