( Ixxviii ) 



with the range of which we are at present ignorant. At any 

 rate, it appears to me inconceivable that any change of 

 environment requiring a modification of structure of suffi- 

 cient magnitude to rank as diagnostic in the systematic sense, 

 should not also be accompanied by a greater or less amount of 

 physiological readjustment. 



The practical outcome of the foregoing suggestions is the 

 very pertinent question whether morphological are not also 

 physiological species, in the sense of having specific physio- 

 logical activities — if not invariably, at least in many cases.* 

 And out of this arises the farther question whether such 

 physiological difl:erences may not be correlated with external 

 characters which might or might not be considered of diag- 

 nostic value. Such characters might be non-significant from 

 the point of view of direct utility, but, by hypothesis, they 

 could never be harmful. The external and the internal 

 characters are alike under the control of natural selection .f 

 If a physiological modification is necesmrihj accompanied by 

 some harmful outward token it could never survive ; but, on 

 the other hand, there is no reason why physiological modifi- 

 cations should not also be correlated with external useful 

 characters, in which case natural selection could take advan- 

 tage of them. We have here the suggestion that physio- 

 logical variability may be the cause of external visible 

 variability, and on this point entomology has much to 

 contribute. 



* Mr. T. D. A. Cockerell has in a resent paper on "The Bees of the 

 Genus Perdifa, F. Smith " (Proc. Acad. Nat. Sci. Philadelphia, Jan , 1896) 

 come to the conclusion that "the essential distinctions between species are 

 physiological, the morphological ones being onlj- valid for diagnostic purposes 

 just so far as they happen to coincide with the physiological." I only came 

 across this statement after the above was written. I am bound to state that 

 the evidence does not appear to me at present to warrant such an extreme 

 view. There must be so much in common in the physiological processes of 

 allied species, that well-marked physiological differences cannot, without 

 further evidence, be regarded as the universal characteristic of sped 5c 

 differences. 



t " With cattle, susceptibility to the attacks of flies is correlated with 

 colour, as is the liability to be poisoned by certain plants ; so that even 

 colour would be thus subjected to the action of natural selection." 

 (" Origin of Species," 6th ed., p. 159. Also the '' Variation of Animals and 

 Plants, etc.," Vol. IT., p. 229.) 



