77 



readjustment." Cockerell, in 1896, wrote. " The essential 

 distinctions between species are physiological, the morpho- 

 logical ones being only valid for diagnostic purposes just so 

 far as they happen to coincide with the physiological." 



Close stud}- of certain groups of Lcpidoptera in which 

 evolution is at present at work in the direction of species 

 formation — Antlivoccra, Ercbia, and Melitcea — incline me 

 strongly to accept Professor Cockerell's view with very 

 slight limitations. It appears to me certain that there is an 

 essential difference between the development of morpho- 

 logical characters and physiological characters, the former 

 tending to maintain the species in statu quo as a species, 

 whilst its colour or superficial appearance is altered, e. g. 

 Gnophos (in many species), Amphidasys bctularia, &c, whilst 

 the latter sets up at once a distinct isolation between the 

 form and the parent stock, preventing any general ten- 

 dency to cross-breeding, and eliminating the progeny when 

 it occurs. Thus the .two extreme forms of Amphidasys 

 bctularia, in spite of their external differences, cross freely in 

 a state of nature; whilst Antlivoccra tvifohi and A. loniccrcc 

 (though equally able to cross and produce fertile progeny) 

 are probably specialised in nature to some particular food- 

 plant, habit, and habitat, and are maintained distinct. 

 Again, A nthroccra palustris appears to be maintained in nature 

 as distinct from A. trifolii, owing to its specialisation in the 

 larval stage to Lotus uligiwsus, the localisation of its. food- 

 plant to a marshy habitat restricting the species to the 

 same habitat, the latter reacting on the species by making it 

 later in its time of appearance than is A . trifolii. There 

 must be some physiological peculiarity, some difference in 

 the physiological processes between two allied species, that 

 limits them to different food-plants, and allows them to die 

 rather than to live on an)' other. Twelve months ago I 

 pointed out how specialisation to food-plant, specialisation 

 by habit, specialisation in the method of hybernation, &c, 

 also resulting from a difference of physiological function, 

 would insure isolation of any particular form, and cut it off 

 from any connection with its nearest allies ; and whilst quite 

 agreeing with Professor Meldolathat " any change of environ- 

 ment requiring a modification of structure of sufficient 

 magnitude to rank as diagnostic in the systematic sense " 

 would also " be accompanied by a greater or less amount of 

 physiological readjustment," yet at the same time one can 

 understand selection to work on some variable factor in the 

 organisation of an insect in such a way as to allow it to take 



