INFLUENCE OF ENVIRONMENT ON SEX 243 



Food 



In view of the work which has from time to time pm'ported to 

 show that nutrition has an influence on sex, an experiment was 

 performed to show whether a difference in amount of food ma- 

 terial might be responsible for the difference of sexual develop- 

 ment of small Crepidulas. The experiment is based on the 

 assumption that the Crepidulas take the same sort of food as the 

 hermit crab, and that they do in fact live on the fragments thrown 

 free in the water about the hermit's crab's head as he tears up 

 food with his chelae and jaws. 



Experiment 8. Carried on in the laboratory aquarium, Woods Hole, 

 Table 12. Small neuter individuals were marked by a notch in the 

 front edge of the shell and transferred to hermit shells from which all 

 but the large females had been removed. All the hydroids were first 

 scraped from the hermit shells in order to remove all possible sources 

 of food about the Crepidulas. The new colonies were then placed in 

 a glass aquarium about 1 foot in diameter and the same in depth, into 

 ■which ran sea water which had first been passed through a Berkefeld 

 filter. After periods specified in the table, the animals were fixed 

 and sectioned. The position in the colony with reference to the 

 females at the end of the experiment is noted in the table. The hermit 

 crabs though unfed remained active and apparently unaffected. 



Table 13. Other specimens from the same lot as those used above 

 were notched and transferred to hermit shells from which all but the 

 large female Crepidulas had been removed, as had been done with the 

 others, and the new "colonies were put in a similar glass aquarium. Sea 

 water which had not been filtered was allowed to run slowly into the 

 jar, and the hermit crabs were fed every day with crushed spider crabs^ 

 small fish, etc., large quantities of which they devoured. Specimens 

 from the colonies were fixed and sectioned at the same times as from 

 the starved colonies. 



It will be observed that the individuals which show least amount 

 of male development or none (specimens 362, 363, 368) had settled 

 themselves after attachment, at some distance from the females. 



The foregoing indicates- that male development is not a mat- 

 ter of food or the lack of it ; for development takes place equally 

 well whether the neuters have plenty of food or none at all, 

 provided only they are near the larger animals. 



A similar experiment, not tabulated here, in which males were 

 used instead of neuters, indicated clearly that starving or feed- 



