56 Riley Presidential Address. 



ished one-half or more, or fully doubled, from the normal, by 

 limiting or increasing the supply of food, as I have proved with 

 Pelopaeus. 



But when we come to the facts in the economy of the Termites, 

 this explanation does not hold good to the same degree. Here 

 we find still greater diversity in form than even among ants, 

 under circumstances where control of these forms by the colony 

 itself must be much less, but nevertheless does occur. The 

 young Termite is to a limited extent, and during early life only, 

 provided with food by members of the colony, and from birth is 

 essentially a free moving agent, less dependent on the adults. 

 We have much yet to learn as to the actual facts, which would 

 seem also to vary in different species. Thus in Eutermes Mr. 

 Hubbard believes, but I think wrongfully, that the young feed 

 on nodules, specially prepared, of comminuted and doubtless 

 partly digested material, while Fritz Mu'ller belie v 7 es that they 

 feed on a fungus mycelium which develops on such prepared 

 substance. The truth with most species seems to be that they 

 are fed on a semi-liquid fluid disgorged from the mouth, whether 

 of -the workers or the undeveloped queens; while in some cases 

 they are fed from a secretion from the anus. (See Note 6.) In 

 these respects and in the early helplessness of the larvae, they 

 closely approximate the social Hymenoptera. 



Similar variations to those found in social insects, whether 

 sexual or seasonal, are extremely common among insects which 

 are not social, as is well exemplified by the long category of phy- 

 tophagic variation, secondary sexual characters, and of dimorph 

 ism and heteromorphism among insects. These variations in non- 

 social insects are often equally as marked and as curious, struc 

 turally, as they are among social species. They are also, except, 

 perhaps, the secondary sexual characters and the- variations which 

 take on the form of mimicry, equally difficult to explain on any 

 view of natural selection that is all-sufficient. On the whole, 

 then, it may safely be said that the production of neuter insects 

 is determined in each generation by the colony itself, in the man 

 ner in which the larvae are fed and reared. In so far as this is 

 true, it is outside the domain of natural selection, and speaks 

 eloquently in favor of the various other causes of variation and 

 modification which have been insisted upon by many of our lead 

 ing American biologists, and which I have repeatedly urged in 



