XXXIX 



would have been no reason, why these peculiarities should be retained by their 

 descendents. The assumption of sudden changes would remove this difficulty 

 at once; but how would this explain the sudden resemblance? However, so 

 many discoveries have been made in this field, that a few more might easily 

 be added ! 



There is one biological phenomenon that is particularly prominent in the 

 Lycaenidae ; unfortunately it is only since I left Java, that attention has been 

 drawn to it, so that, being ignorant of it, I have not made any observation 

 upon it. I refer to the symbiosis, which occurs in this family in so many 

 forms, that it can hardly be attributed to one origin, which makes the explanation 

 of it very difficult. It is now known, that the larvae of many Lycaenidae are 

 provided with glands which secrete a sweet liquid, in the same way as the 

 Aphididae, and, also as in the case of the Aphididae, that this is collected by 

 ants, who use them as milch- cows, stroking them with their antennae, to 

 encourage the secretion. There are many publications on the subject, including 

 such as maintain, that in return for this service the ants protected the larvae against 

 their enemies, and that they drive them to their nests to breed them. Not long 

 ago I read an article about one of the Lycaenidae-caterpillar in East Africa, that is 

 supposed to live inside a Cecidion and be fed there with pieces of leaves by the 

 ants. They are also supposed to take care of the pupae, and to obtain tissues for 

 their nests make use of the larvae as spools, an account of their spinning 

 capacities. It is doubtful whether a good deal of fancy and auto-suggestion 

 are not conducive to such observations; at least experience teaches us to be 

 cautious in accepting them. Nevertheless the fact of symbiosis is indubitable. 

 I have not, as I say, observed it myself Although I had noticed in my 

 garden in Batavia, that the numerous larvae of Lycaena Panda va Horsf. that 

 I found there on a species of Cycas, were always surrounded by a swarm of 

 ants. I was surprised to find that they apparently did them no harm, and being 

 at that time still obsessed by the theory of mimicry, I imagined, that they 

 must be protected by some sort of mimicry and tried to discover the nature 

 of this. Naturally this was attended with no result, except the negative one 

 that, concentrating all my attention upon this point, I failed to grasp the real 

 nature of this symbiosis. 



Subsequently Edw. Jacobson, being acquainted with the phenomenon, 

 observed it with the Javanese Hypolycaena Erylus Gdt. and the ant Oeco- 

 PHYLLA Smaragdina, publishing his results in part LV of the Tijdschtift voor 



