PIGMENTATION OF A POLYCLAD. 727 



It is of more value to discover the source of the pigmentation than 

 to speculate concerning its "function." A simple experiment showed 

 that, as was to be expected from what is known of some other turbel- 

 larians, 4 a good fraction of the color seen in such specimens as those 

 drawn in Figures 2 and 7 was due directly to food in the alimentary 

 spaces. 



The individual shown in Figures 2 ami 3 was isolated from its 

 Ectinascidia colony (cf. Fig. 1) on July 10, 191G. Five hours later 

 it had become considerably paler (Fig. 4) ; and two days later it was 

 very conspicuously so, as indicated in Figures 5 and 6. It was then 

 returned to a small dish containing several Fctinascidia zooids, one of 

 which the flatworm very promptly found and began to feed upon, 

 creeping over its surface. After six hours it had assumed a brilliant 

 orange color, like that of Figure 2. The major portion of the orange 

 pigment comes, then, directly from the excreted pigment-bearing cells 

 at the. surface of the tunicate test. A similar experiment was made, 

 with a corresponding result, in the case of one of the purplish-black 

 Hatworms found with A. atra. 5 



There appears, however, to be a minimum below which the pigmen- 

 tation cannot be reduced by moderate starvation. After four days 

 in seawater, removed from their host, two of the orange specimens 

 were still in the condition depicted by Figure 5. So far as could be 

 made out from microscopic examination under gentle compression, 

 no orange pigment was retained in the digestive system, although some 

 was present in the integument. 



The readiness with which these polyclads return to their own partic- 

 ular kind of tunicate is surprising. With all of the individuals ob- 

 tained, tests were made by giving them the opportunity of feeding upon 

 E. tubinata or A. atra, placed together in the aquarium, and in two 

 cases specimens found on E. turbinata were also offered individuals of 

 A. curvata. When kept in aquaria the polyclads frequently left their 

 host and wandered about the dish, but in all these instances, as well 

 as in numerous trials in which they were artificially removed, they 

 always returned to that type of ascidian upon which they were origi- 

 nally found. Nor could they be induced to feed upon a different species. 

 I tried in this way to alter the coloration of the specimen found in 

 A. curvata, and of two of those taken on Ectinascidia, namely by plac- 



4 It should be stated, though, that practically no observations have been 

 recorded respecting the food of the brilliantly colored polyclads. 



5 For a note on this pigment, see Crozier (1916). 



