78 



up a comparatively vacant place in the economy of nature ; 

 whilst, at the same time, in other directions selection seizes 

 on other external morphological characters, and fits by their 

 development the insect still more completely for the changed 

 conditions in which it finds itself, — that is to say, the 

 primary conditions producing the change may be physio- 

 logical, adjustment to the environment following as a matter 

 of course. 



I am insistmg on this point because I wish to engage the 

 attention if possible of some of our field naturalists to the 

 points at issue. We want to know absolutely whether, in 

 nature, certain species or forms are specialised by their 

 attachment to certain food-plants, or by certain habits that 

 differ from those of their nearest allies. This is true 

 naturalist's work, work for the observer in the field, and not 

 to be discovered by the examination of cabinet specimens. 

 It is quite evident if internal structures and internal func- 

 tions are variable, and that natural selection can act on 

 these in the same way that it acts on variations of external 

 structure, when such modifications are required in response 

 to change in external conditions, that we have here factors 

 of isolation in forming species that need have but little in- 

 dication of their distinctness in external character and form. 

 It is in this sense that Tephrosia crepicscularia and T. histortnta 

 are species, their isolation having been begotten by a differ- 

 ence of physiological organisation, a difference that insists 

 on the one form, T. bistortata, appearing in the spring some 

 weeks earlier than the other, and tends to make it double- 

 brooded, whilst the other, T. crepuscularia, appears later, and 

 is religiously single-brooded. 



Is difference of habit such as this a sufficient criterion of 

 a difference of species? Meldola, speaking of characters 

 important to a species that may be altogether left out in the 

 diagnosis of cabinet specimens, refers to the disguises adopted 

 by insects in the numerous cases of protective resemblance, 

 and the mimicry that are so familiar to all of us, and adds, 

 " If utility is not obvious in all such instances, then nothing 

 in the realm of organic nature will bear the interpretation 

 of utility. . . . Now these adaptive and demonstrably 

 useful characters are surely 'specific,' whether the syste- 

 matist attaches much or little weight to them in his diagnoses. 

 Moreover the disguise is enhanced, and in many cases is 

 only really effective, when combined with certain habits 

 which are not and cannot be taken into consideration in 

 ordinary diagnostic work. Nevertheless such habits are as 



