160 MR. R. MELDOLA ON VARIABLE PROTECTIVE [Feb. 4, 



duct of natural selection, which works in this case, as in all other 

 cases, upon the variations of colour presented to it by nature, quite 

 regardless of the manner in which the colour is primarily produced 

 — regardless whether it is hypodermal or epidermal, whether it is a 

 colour due to interference, or a colour due to pigment-cells in the 

 skin. 



A similar mode of reasoning applied to such cases of variable 

 protective colouring as have been considered in this paper will, I 

 imagine, serve to establish the truth of the proposition, that such 

 cases are " attributable, in great part at least, to the action of the 

 ' survival of the fittest.' " 



Let us take the several groups of cases included in Class V. in 

 the order already dealt with, and apply this reasoning to them. We 

 have first to deal with larvae wliich feed on several plants of different 

 colours, and which are capable of adapting themselves to the colour 

 of the particular plant on which they are feeding. Now, granting 

 that this power of changing colour is beneficial to the insects by 

 affording them concealment, a truth which no entomologist who has 

 witnessed any of these cases will deny, it follows that natural selec- 

 tion would eliminate any variations tending to depart from this 

 useful power of adaptability to the colour of the food-plant. Here, 

 then, the function of natural selection, as in the illustration first 

 brought forward, is simply to maintain a power possessed by the 

 larva, regardless whether this power resulted in the first instance 

 from the direct action of external conditions — regardless whether it 

 is under the control of the creature's will or not. Assuming, in 

 these cases, that the change of colour is due to the presence of the 

 colouring-matter of the food-plant in the tissues of the caterpillar 

 (as it most probably is), we might say in more concrete language that 

 natural selection is and has beeu at work weeding out those indivi- 

 duals whose skins were not sufficiently transparent to allow the 

 colouring-matter to appear through them. 



Passing on to those pupae which appear to be photographically 

 sensitive, we shall find a similar mode of reasoning to obtain. 

 Larvae in selecting a suitable place to assume the pupal condition 

 are liable to be exposed on surfaces of different colours. It will be 

 admitted that pupae which harmonize with the colour of their rest- 

 ing-surface would be more likely to escape detection than individuals 

 not thus coloured. I need only allude, en passant, to the perfect 

 manner in which the pupa of Synchloe brassicce matches the speckled 

 wall on which we so often find it. Now a pupa is liable to be ex- 

 posed on a surface of any colour ; how can such a state of affairs 

 be met? obviously only by giving to each individual a power of 

 changing colour in correspondence with the colour of its resting- 

 surface. Observe, now, that we are not here in any way concerned 

 with the primary cause of such a faculty : natural selection only 

 takes advantage of the property, no matter how it has originated. 



We may finally proceed to the examples of this class furnished 

 by insects in the perfect state. An insect adapted to the colour of 

 one district, but forced to roam into other districts in search of food 



