STUDIES OF VARIATION IN INSECTS 211 



nearly exactly uniform as could be guaranteed under our most 

 careful artificial (experimental) conditions. The pupae are, as 

 well, under identical conditions of temperature, moisture and 

 light, so that when the adults issue the variations to be found in 

 any of their parts may with complete confidence be ascribed to 

 prenatal influences, to intrinsic causes. They are purely blas- 

 togenic. Similarly, the conditions of life of the developing in- 

 dividuals of all the other social insects, the termites, ants and 

 social wasps, are practically identical. 



With the insects with incomplete metamorphosis, as the 

 Orthoptera, Hemiptera, Corrodentia, et al. (that is, those which 

 hatch from the o.^^ in the general guise of the adult, although 

 always wingless and of minute size) the appendages of the 

 head (mouth parts and antennae), as well as the eyes and general 

 head wall and also the legs and external genitalia, are the same 

 organs as the corresponding adult ones ; whatever developmental 

 changes take place are of the nature of a gradual modification 

 from immature to definitive imaginal condition. The wings 

 however, are never functional until after the final moulting (the 

 one just prior to the assumption of the complete imaginal condi- 

 tion of the whole body) and so are probably not capable of modi- 

 fication (adaptive variation) due to use or disuse or to any 

 direct special influence of the environment. Variations in wing 

 characters of adult individuals of any insect species of incom- 

 plete metamorphosis ought therefore to be strictly congenital 

 (other than differences, as of size, perhaps, due to the influence 

 of nutrition, humidity, temperature, and light). But such char- 

 acters as the color and pattern of the body-wall might very well 

 be modified by the direct influence of the environment. For 

 example with certain genera of locusts, as Trimcrotropisy 

 SpharagemoUy et al., there are marked individual variations in 

 coloration and pattern, obviously adaptive for the sake of pro- 

 tective resemblance. It may well happen that the individuals 

 of a single brood starting life of similar color and pattern should 

 acquire varying protective coloration modifications, if such ac- 

 quirement be actually possible. ' We know from the experiments 

 of Poulton and others that the body-wall of certain insects, par- 

 ticularly the larvae of Papilio butterflies, can respond directly 



