XXYIU. INTRODUCTION. 



it is probably also an occasional babit with many others, that the 

 female will attack and kill, and even devour the male ; now 

 it is obvious that in her attempts to catch him, the chance of 

 escape will be in proportion to the smallness of his size and his 

 activity ; thus the larger the male (it being remombered, of course, 

 that the male is always, excepting in rare instances, the 

 smaller and weaker of the two) the greater his risk of being 

 destroyed, instead of becoming the progenitor of a future race. 

 The smaller males will therefore more commonly be the ones to 

 perpetuate themselves; and so, by a kind of "natural selec- 

 tion " this sex will become gradually less and less in size, until 

 that descending limit is reached beyond which the male would 

 not be large enough to perform the offices for which it was 

 created. That limit certainly seems to have been reached in the 

 case of the tropical spider above alluded to. The disparity 

 in size between the male and female of JSpeira diademata is also 

 very great. 



It should be remarked that in measuring spiders the " length" 

 is, unless otherwiso stated, the length of the cephalo-thorax and 

 abdomen, exclusive of the falces. 



Protective resemblance of Spiders to Inseots or other 

 Objects. 



Another very interesting part of the study of spiders (as 

 indeed it is of all the Aktictjlata) is the way in which some 

 species resemble others, belonging often to very widely 

 separated groups of this branch of the Animal Kingdom. This 

 resemblance has been called "mimicry;" an unfoitunate term 

 as it seems to me, because it appears to imply conscious imita. 

 tion, in which, of course, no naturalist (so far as I am aware) 

 believes for a moment. The resemblance is eometimos exceed- 

 ingly close, and may be called mimetic ; but the more abstrac^ 

 phrase, " protective resemblance," expresses all that is really 

 understood by the fact, and does not seem to be liable to miscon. 

 struction. The resemblance is undoubted, and its protective effect 

 has been proved in some cases, and may be assumed in almost 

 all. The most striking instance of "resemblance" known to 



