ZooL— Vol. III.] UBANCROFT— COMPOUND ASCIDIANS. 1 57 



Differences in color were observed as follows: — 



(a) In the same zooid at different ages. 



(b) In the same colony, resulting from age, vigor of 

 growth, and changes of environment. 



(c) Between the various colonies derived from the same 

 parent, and subjected to the same external conditions. 



In the previous section the nature of the color variation, 

 as such, has been considered. Here we are concerned 

 only with the problem whether it is permissible to use 

 differences in color in Botrylhis as diagnostic characters. 

 Certainly differences belonging to any of the three classes 

 mentioned above maybe affirmed to be of no specific value; 

 and accordingly the plan will be to compare the colors of 

 the zooids observed at different times in the same colony, 

 and in brother and sister colonies, with the colors of the 

 various species of Botrylhis described. 



Before beginning this comparison, however, it should be 

 again pointed out that only a small part of the variations 

 obtained are recorded in the figures. Many of the colonies 

 died or were killed by some of their many enemies before 

 they could be figured. The color also changed so fre- 

 quently that it was impossible to record all the variations. 

 Furthermore, the zooids figured were not selected with a 

 view to their comparison with described species, but rather 

 to record the amount of variation, and therefore many 

 zooids that had very nearly the same color as some of the 

 described species were not figured. 



In spite of these disadvantages, however, the figures of 

 the members of these three families show quite a close 

 approximation to seven described species.^ In Family I 

 there are: — 



Botrylhis aiiroUneatiis Giard, var. t Della Valle, 1877, 

 p. 28. 

 Figure 8 corresponds closely to the description, except 

 that the coating of white may extend too far anteriorly, 



1 Unfortunately, the original figures have not been accurately reproduced in all cases, 

 and consequently the originals will be gladly sent to any one especially interested in 

 the subject. The extent of the variation, however, is quite accurately represented. 



