168 THE WORLD OF LIFE 



CHAP. 



processes of nature by the help of facts known to us. I can 

 say this with more confidence, because I find that I have 

 myself come to a hasty conclusion, which I now see to be 

 erroneous, on one aspect of this very question ; and as it in- 

 volves a problem of some importance I will here state what it is. 

 I find that in all my writings on this subject I have assumed, 

 without going into details, that the theory of " recognition- 

 marks," which so well accounts for a very widespread type 

 of marking and coloration in birds and mammals, is also 

 applicable to a large portion of the markings of insects, 

 especially in the case of butterflies. But a little consideration 

 shows that there is no resemblance between the two cases. 

 Young mammals and birds grow up with their parents, and 

 get to know their appearance in every detail. They also 

 have usually brothers and sisters growing up with them, so 

 that by the time they go out into the world to care for them- 

 selves they are thoroughly acquainted with the difference 

 between themselves and other species, even those nearly 

 allied to them. This complete knowledge is increased by 

 the fact that they are able, through the mobility of the head 

 and neck, to see almost every part of their own bodies, and 

 thus know that they themselves do resemble their parents. 



But with the butterflies, and most other insects, every- 

 thing is different. The caterpillar never knows its parent, 

 and when the butterfly emerges from the pupa and takes 

 flight, it seems quite impossible that, among the numerous 

 butterflies of all sizes, shapes, and colours that it may 

 immediately encounter, it can possibly know, by sight, which 

 are of its own race. It must be remembered that from the 

 position of its eyes it cannot see itself except at so oblique 

 an angle as to be almost useless ; and when we consider the 

 extreme diversity of the sexes in many butterflies this adds 

 to the difficulty of supposing vision to be the primary means 

 of recognition. But it may be a secondary means. It is 

 well known that in some moths the females attract males 

 by scores at night, and this can only be by scent, or some- 

 thing analogous to it. It is also known that the males of 

 many butterflies emit a strong perfume which has been 

 traced to certain peculiarly formed scales on the wings. 

 Scales, apparently of a similar nature, have been found in 



