61 



ately mimicking leaf-form, and with the dorsal saddle ranging in 

 colour from purple or purple-grey to cream. 



(We have not taken the Kittens in sufficient quantity to learn 

 if their pattern is variable). 



In the full grown larvae the mimetic effect of the colour 

 scheme is heightened by the light border which separates the green 

 from the purplish areas, the thin light line picking out the design 

 in accurate mimicry of high lights caught by thin leaf edges, or 

 by the edges of holes perforated in leaf tissue. This last effect 

 is seen to perfection in larvae which have the cusps of the dorsal 

 pattern as they approach the ventral line detached from the main 

 saddle, when the islands of purple, margined with white, irre- 

 sistibly suggest holes in green with shadow coming through. 



The Puss at maturity hanging as is its wont back downward 

 from petiole or leaf edge presents in an oblique position when 

 attached to its support at a high angle, the suggestion, when seen 

 laterally as a passing bird might espy it, of two green leaves 

 against a background of purple bark or shade — while in a hori- 

 zontal position — which on the average of occurrences in the 

 foliage of a large Populus must be the commonest posture — the 

 dorsal saddle is an accurate copy of the slightly upturned tip of 

 a fore-shortened leaf. So good is either resemblance that work- 

 ing men shown the full grown larvae at two feet distance failed 

 to detect them, or recognise a caterpillar till convinced by the 

 evidence of touch. 



Whence has the light marginal line which lends all its 

 significance to the pattern originated? 



To answer this question we must consider the relative survival- 

 value to the race of the growing larva at each moult. Suppose 

 the Puss Moth to lay two hundred eggs — a modest estimate — 

 to keep up the supply of the species, year by year, only two of 

 these need complete the life cycle. Regarding the successive life 

 stages as heats in a race, we have five heats, and if as improbable 

 we may dismiss the pupal dangers as relatively negligible, we 

 shall be well within the mark in estimating a loss of half the com- 

 petitors in each heat. The survivors in the last heat have evidently 

 a survival-value equal to that of the whole batch at hatching ; 

 they are diminished to one sixteenth of their original total and 

 each individual has risen sixteen fold in survival-value — they are 

 large and toothsome and urgently in need of concealment. The 

 light line now so effective a detail of the colour scheme would be, 

 if broad, a danger rather than an element of safety. This there- 

 fore precludes its earlier development. Can we hazard the con- 

 jecture that the ancestral puss caterpillar lived in its last stage, as 

 its descendants still do in the four earlier moults, on the upper 

 side of the leafage — that it was smaller, and that the wavy light 

 line was then broader and in all the stages of functional value? 



A leaf, seen from above as a passing bird would see it, if 

 injured presents the following appearances : — 



1. A mere scratch or mechanical lesion will give a pale mark 



