360 AMERICAN SPIDERS AND THEIR SPINNINGWORK. 



young stay with the mother until they are somewhat grown, and then 

 have shelter beneath rocks or other secluded positions where birds are not 

 apt to find them, although, of course, some birds do mouse around and 

 pick up insects in the most secluded spots. I imagine, however, that the 

 dangers threatening young Attidae are much greater from certain insects 

 and individuals of their own order, than from birds. I can, therefore, 

 hardly conceive what advantage it would be under such circumstances for 

 the spider to resemble an ant, even if we were to suppose that such a 

 minute resemblance as the hypothesis requires at the outset, would be of 

 advantage in any case. In point of fact, the theory is not workable, as it 

 seems to me. Any change of form to be effective must occur in the first 

 stages of life. But a minute resemblance could be of no advantage, as the 

 discriminating powers of enemies, whether insects, birds, or spiders, are 

 hardly so acute and delicate as to make an infinitesimal variation of 

 much importance in screening one individual spiderling in the rnidst of 

 a brood. In short, if we are to suppose that the birds are the real ene- 

 mies, or any other creature that is indisposed to feed upon ants, it seems 

 necessary, in order to justify anything like this theory of the origin of 

 mimicry, to suppose that the variation of the spiderling was, at the outset, 

 so great as to give it at least a reasonable likeness to the ant. 



The theory takes for granted an accuracy of eyesight on the part of 

 birds that close observers will scarcely be willing to admit. How far can 

 birds distinguish color? This is a question which has scarcely 

 yet been fully solved. How far can insects distinguish color? 

 How far can birds and insects distinguish between such minute 

 variations in form as that which the above theory seems to require? In 

 accounting for the origin of cocoon mimicry we suppose that the eyesight 

 of birds and wasps is so defective as to form and color as to permit them 

 to be deceived by a difference as little marked as that which exists between 

 the cocoon of the Tailed spider, for example, and the spider herself. But, 

 in accounting for the mimicry of ant like forms, we are compelled to re- 

 verse this attitude, and suppose the eyesight of birds and raiding wasps 

 to be so accurate that it can distinguish between a slight variation on the 

 part of one spider of a brood towards an ant, and the normal form of 

 other spiders of the brood, and distinguish so accurately that it will 

 avoid the ant resembling spiderling and. take others. Whatever theory 

 of the origin of mimicry we adopt, certainly must be free from inconsist- 

 encies such as this. 



Moreover, the greatest destruction, as far as I am aware, wrought by 

 birds upon young spiders is accomplished under circumstances that pre- 

 clude any such an element as above. Immense numbers of spiderlings, 

 including, I believe, all species of Saltigrades, possess the aeronautic habit, 

 and while they are flying through the air upon their tiny mimic balloons 

 they are devoured by swifts and swallows, who skim the air and gather 



