156 CALIFORNIA ACADEMY OF SCIENCES. [Proc. 3D Ser. 



the older colonies. But as the colony grows, each suc- 

 ceeding generation of zooids reaches a greater size than 

 the preceding. The maximum is usually reached while 

 the colony is still young, lightly pigmented, and growing 

 vigorously. The edge of the colony is always growing 

 faster than the center; and when the zooids of the same 

 colony differ in size, it is always at the edge that the 

 largest zooids are found. When the colonies have about 

 reached their maximum size, and their color patterns are 

 well developed, but the growth is a little slower, the maxi- 

 mum size of the zooids is somewhat reduced; and, finally, 

 when the colony is old, and the number of zooids decreas- 

 ing, the latter never grow much larger than when the colony 

 was very young (fig, 16). Figures 17 to 21, all drawn to 

 the same scale, show some of the variations in the size of 

 the adult zooids of the same colony. 



4. Colors of Zooids and Ampulla. 



The colors of the zooids and ampullae have been used so 

 much in distinguishing the described species oi Bolryllus that 

 the other characters just discussed are of very little impor- 

 tance when compared with them. Almost every described 

 species has been partially, and many entirely, based upon 

 color characters. The frequency with which these charac- 

 ters have been used for diagnostic purposes undoubtedly 

 depends upon the immense variation to which they are 

 subject. This is so great that in any lot of Botryllus indis- 

 criminately collected it is usually somewhat difficult to find 

 two with exactly the same colors. The large extent of this 

 variation has often been emphasized. In fact, several of 

 the authors that have described species of Botryllus have 

 expressed doubt as to whether it was really species they 

 were describing and not merely variations. 



The evidence that is here offered as to whether the 

 variations in color furnish characters that are of value in 

 separating species depends upon the observation of the 

 three families of colonies mentioned (pp. 139, 143). 



