PROTECTIVE RESEMBLANCE. 167 



trees and other foodplants are constantly being destroyed in the spread 

 of large building, with here and there an isolated tree left, where a 

 year or two of inbreeding would be sufficient to start a dark race. 



Mr. Newman of Bexley, by inbreeding Kimonios an(/iilaria, succeeded 

 in producing in the third generation, 25% melanic specimens. By 

 inbreeding and selection the Rev. G. H. Raynor has succeeded in 

 breeding melanic and other interesting forms of Ahra.ras (/ro!<fiida)iata, 

 but I have seen most of these forms bred from wild Lancashire larvae, 

 collected in the neighbourhood of a large town. How has this been 

 brought about ? I suggest either inbreeding in nature, or else they 

 are the result of inbreeding carried on by the late C. S. Gregson and 

 others some years ago in their quest for varieties, who not knowing 

 Mendel's Law, turned out the typical examples which laid ova that 

 produced varieties in the following season. If the latter is the true 

 solution then it is interesting as showing that once the varieties are 

 started they will recur after so long a time. 



In British Nuctiiae and Varieties, vol. i., p. 17, the late Mr. Tutt, 

 in speaking of Acmnicta aliii says, " It would appear, from a brood that 

 Dr. Chapman reared, during the summer of 1890, that the dark forms 

 of this species are probably connected with a change of constitution 

 (disease ?) " and then quotes Dr. Chapman (Ent. Record, vol. i., 

 p. 271-2) who Avrites, " I have bred «?)«/ for several years. . . . The 

 carious point, however, in the case of alni is, that in previous years 

 hardly a specimen departed, even in a slight degree, from the normal 

 type, whilst this year, about a third of the specimens difit'er, more or 

 less, either in the suffusion of the pale areas with darker scales, or in 

 variation of the stigmata, generally in the direction of disappearance 

 •of the orbicular one." 



Inbreeding, however, does not always produce melanism. By 

 inbreeding Amorpha jxi/ndi I obtained, by a pairing of typical grey 

 specimens, a most variable series, including pink, fawn, fawn with the 

 usual markings nearly obliterated, violet, etc., specimens. 



Diaitthoecia carpopJuifia at Eastbourne and Folkestone, more or less 

 on the chalk, are pure white or pale fawn and large specimens, whereas 

 at Croydon, also on the chalk they are ochreous, and the two series seen 

 side by side are quite distinct. If the weeding-out by birds had any- 

 thing to do with it, one would expect the same result in all three places. 

 Specimens from Lancashire are smaller and termed melanic. If, 

 however, we compare them side by side with their sister species, D. 

 i-ajisincola, from Eastbourne, we find the colour is much about the 

 same, yet the latter is not usually referred to as melanic. V. capsincola 

 has practically the same habit and habitat at Eastbourne and elsewhere 

 in the south as D. carpajdiana. If the birds had weeded out the dark 

 D. carpophat/a, why have they not also weeded out the dark D. capxincola 

 and " natural selection by means of protective resemblance," produced 

 pale forms. I do not think that " protective resemblance," will solve 

 these colour problems. 



Why is it that D. carpophaf/a and D. conspersa are so variable, and 

 7). irrefptlaris, D. alhiwacula, D. cucubali, 1), capaincola, and D. capso- 

 phila so constant ? 



Why is it that D. cacmi and H. wrena vary much on the Continent 

 and so little in England '? 



Why is Luperjna testacea so variable and f,. ceapitis so constant? 



