10 THE entomologist's KECORD. 



extent of 70 per cent., the majority of the remainder having 5 teeth 

 and the rest 3 or 6. The latter, then, as regards scales, showed a 

 similarity to russata, and no differentiation seems possible. These two 

 last results, based on single specimens, require confirmation or other- 

 wise, but it would appear as though only the one form, marmorata, can 

 be differentiated by its scales. 



Experiments were also made with the two Cuspidia, tridens and pst. 

 I have only as yet been able to examine two specimens of tridens, not 

 caring to sacrifice more of my series of the bred insect. The average 

 of the two gives — 



Scales with 4 teeth formed 35 per cent, of the whole.* 



}) '-' )f }} '-" J5 5) 



100 



or, 65 per cent, were scales with 5 or 6 teeth. One had fewer scales 

 with 6 teeth and more with 4 than the other, but there was the pre- 

 dominance of those with 5, in both. Psi on the other hand showed — 



Scales with 3 teeth formed 6 per cent, of the whole.* 

 4 74 



100 

 or, the scales with 3 or 4 teeth formed 80 per cent. 



Not only is there this difference between the two, but the scales of 

 tridens have also unequal and irregular teeth with projection of the 

 middle ones, giving a ragged appearance, in a very considerable pro- 

 portion, especially of those from the centre and hind margin of the 

 wings, whilst the teeth of the scales of jysi are comparatively equal and 

 regular, and show less tendency to projection of the middle ones. 

 This difference is very noticeable. Of course these points remain oj^en 

 for confirmation or otherwise when more bred sjjecimens of tridens 

 and the darker varieties of psi have been examined. As regards any 

 connection between colour and intensity of colour and the number of 

 teeth of scales, I think, for the most part, the darker insects have their 

 scales with the most teeth, but this is far from invariable, as is seen 

 above, those from the white centre of var. marmorata having more teeth 

 than those from the true dark imvianata or the var. per/uscata of russata, 

 and I have noticed other similar cases. 



I wish these notes to be taken as suggestive for the most part, 

 though I have endeavoured to some extent to guard against the fallacy 

 of too few data by examining the scales from three different parts of 

 each wing and from a considerable number of specimens (except in the 

 cases of immanata and tridens), still I am quite aware of the fact that 

 many more examinations, corroborative or otherwise, are needed before 

 attempting to generalize, but I think, at all events, I have made out a 

 case for further investigation. — Buckerell Lodge, near Honiton. 

 November 2Uh, 1893. 



*0n the other hand, the total of scales with 4 and 5 teeth comhined give 

 respectively 91 and 92 percent., a very close result. — Ed. 



