ZOOL.-VOL. III.] BANCROFT-COMPOUND ASCIDIANS. l6l 



shape of these markings is usually very variable; for 

 example, at the time when figs. 20 and 21 were drawn 

 many of the zooids in the colony had complete dorsal 

 bands. Figures 18 and 19 offer another example. The 

 color, however, is often remarkably constant. For instance, 

 the white color of the dorsal bands remains constant for all 

 the members of Family 11. In the colony whose zooids 

 are represented in fig. 8, however, the mantle later as- 

 sumed a lemon-yellow and in some zooids even a decided 

 orange color. 



5. The presence of a color pattern developed about the 

 branchial orifice (figs, i, 8, 11, 26, 27) or on other parts of 

 the zooid (figs. 10, 21, 22) is of no value, for this varies 

 greatly in the different colonies of a family or even in the 

 same colony at different periods. Patterns of this kind, 

 that depend upon the unequal distribution of the two kinds 

 of pigment making up the ground-color (figs. 8, 11, 21, 23), 

 were always seen to disappear when the colony was placed 

 in unfavorable conditions (figs. 2, 4) or lost its vigor through 

 age (fig. 16). 



The fist given includes about all of the possible color 

 characters used for diagnostic purposes; and it points 

 inevitably to the -conclusion that in Botryllus, as it occurs in 

 Europe and the Atlantic Coast of North America,^ color 

 characters cannot be used for separating sfecies; and that, 

 therefore, since none of the described sfecies have been based 

 upon morphological characters, there is no valid reason for 

 recognizing more than the single species, B. schlosseri 

 (Pallas, 1766, pp. 355-356) Savigny (1816). 



It might be thought that B. gouldii Verrill (1871, p. 211) 

 of the Atlantic Coast of North America should be excepted 

 in this general lumping of species; but so far as can be 

 seen at present, there is no ground for making an excep- 

 tion, for Botryllus at Woods Hole and at Newport exactly 



1 The species on the Pacific Coast of North America {B. magmas Ritter (Proc. Wash- 

 ington Acad. Sci.. Vol. HI, p. 255), and an undescribed species) do not show any color 

 variation, and appear structurally distinct from B. schlossert. 



