The Bionomics of South African Insects. 375 



at present explain. As a matter of fact there seem to be 

 really very few "blues" in which the tails bear any real 

 resemblance to antennae. Again, so far as the special 

 explanation of jaws is concerned it seems to me that this 

 would rather prompt a bird or lizard to attack the insect 

 at the other end, which would be fatal. In some cases 

 there appears to be a possibility of explaining the particular 

 shape of a tail ; for example, in Gharaxcs the general rule is 

 two thin tails on each wing, and we can understand that 

 this would not be suitable for the leaf-like under-side of 

 C. varancs, which has consequently developed a single 

 thick tail which is more in keeping with its style of 

 coloration. Again, the thick twisted tail of Myrina greatly 

 enhances its general resemblance to a bit of shrivelled 

 fig-leaf, and so forth." 



I think, however, that it is probable that such resem- 

 blance as there is to a head, in certain species of Lyavnidie 

 may be of value and may have been producefl by direct 

 selective action, and I would specially draw attention to 

 Mr. Floyer's suggestion (p. 374) that the butterfly may 

 dart off in a direction which the head-like appearance has 

 caused to be unexpected by an enemy. Many years ago 

 my friend Dr. A. C. Haddon, F.R.S., showed me a specimen 

 of a little yellow fish, about 1\ inches long, which he had 

 observed and captured among coral, Aug, 11, 1888, at 

 Thursday Island, Torres Straits. The head was crossed 

 by a dark, white-bordered, vertical, somewhat curved band, 

 which included the eye and tended to conceal it. At the 

 root of the tail was a very conspicuous eye-like mark. 

 The fish had the habit of often swimming for a little 

 distance very slowly tail first, but if disturbed it would 

 dart off with great rapidity in the opposite direction, viz. 

 head first. That so similar an adaptation should be met 

 with in such a very different part of the animal kingdom 

 affords considerable indirect support to the interpretation 

 of these Lyca^nid marks and structures, at which so many 

 naturalists have independently arrived. Dr. Haddon 

 kindly permits me to make use of his interesting observa- 

 tion, which has not been hitherto recorded, Mr, G. A, 

 Boulenger, F.R.S., informs me that the fish is ChcVtodon 

 lileleius. 



