in Temperature- Variation. 44.1 
Araschnia levana. Cooling tends in the first place to 
separate certain constituents of the ocellus, and when 
carried to a high extent has the remarkable effect of 
causing an unmistakable resolution of the ocellus in the 
forewing, the appearance finally produced being that of 
the ordinary Vanessa character in a comparatively un- 
modified form. It is interesting to see how completely 
these cooled specimens bear out the views which I ven- 
tured to express, some years ago, on the origin and 
constitution of this ocellus (Trans. Hnt. Soc. Lond., 
1390; pp: 99, 100, pl. i., fig. 12). The ocellus of the 
hindwing is also affected in the same direction. 
III. Vanessa POLYCHLOROS. 
Cooling tends to produce several features which 
appear to be ancestral. The chief of these are (1) the 
pupilling with black of the spots of Series D in the fore- 
wing; (2) the occasional indication of Series III. in the 
hindwing; and (3) the tendency towards the formation 
of a new dark spot between II.8 and III.8. These 
points approximate towards the condition in Grapta. 
IV. Grapra C-ALBUM. 
In both broods cooling tends to induce or increase a 
darkness of ground-colour; this being undoubtedly an 
ancestral character.* 
Tn all cases of this kind the obvious question occurs— 
are we to consider these phenomena as true instances of 
reversion, or is it merely that like causes have produced 
de novo a like effect in descendant and ancestor? ‘The 
latter explanation may account for some of the facts, but, 
I think, not for all. It may perhaps give the reason for 
a general diffusion of bluish scales, or for a change of the 
ground-colour from black to brown, but it is scarcely 
adequate to explain the special formation of a definite 
pattern, as of Series III. with its blue centres in JV. 
atalanta, or the reduction of the ocellus in JV. io to the 
primitive Vanessa condition. Without raising the vexed 
question of sexual selection, we may yet affirm that 
* The observations on the three last species are new ; those on 
V. atalanta, as has been seen, are partly new and partly old. On 
the whole subject of the ancestral markings in Argynnis and 
Vanessa, see the author’s paper in Trans. Ent. Soc. Lond., 1890. 
