THE PHILOSOPHY OF XESURUS 287 



which might be operative in connection with it, and 

 I obtained decidedly positive results. Other more 

 elaborate experiments had to be abandoned during 

 this expedition. I took a live Xesurus and armed 

 with thick gloves I bent its tail slightly around 

 and rasped the sharp files against the scales of 

 three species of fish, one a much larger form, 

 Seriola dorsalis, and two smaller than the surgeon- 

 fish, a Poinacentnis arcifrons, and an Evoplites 

 viridis, both of which live in the same locality as 

 the Xeswrus. I had no large carnivorous fish, 

 but it is unlikely that the results would have been 

 different. In each case I had a number of other 

 individuals of the same species, as controls, all 

 living well in our aquariums. I watched the fish 

 carefully but after the excitement, due to my tak- 

 ing them from the tank, was over, I saw no symp- 

 toms of discomfort, the abrasions themselves being 

 quite negligible. The following morning all of 

 the four subjects of the experiment were dead, their 

 fellows, without exception, being still in perfect 

 health. There was a slight discoloration of the flesh 

 about the rasjDed wound, but no other lesions. 



In the case of a butterfly protected by nauseous 

 juices, every inexperienced bird and lizard has 

 probably to catch and taste for himself — the race 

 of butterflies winning immunity at the sacrifice of 

 one of their number. Turning to the life and 

 death problem of Xemrus, from a general point of 

 view there seem to be only three methods of corre- 

 lating the various possibilities and factors. Cor- 

 responding with the case of the butterfly and the 



