SEASONAL POLYMORPHISM. 173 



specimen of either race could be mixed up with the other, such as is 

 not often the case between races. The difference in size gives the 

 impression that the Tyrolese race is gigantic ; actual measurements 

 between the wing tips give 29, 31 and 83 mm. as minimum, usual, and 

 maximum expanse in male, whereas the corresponding measure- 

 in the nymotypical race are 25, 27 and 29 ; the difference 

 between the few females I possess is not as marked, because 

 the ones from Tyrol are smaller than their males. The tone 

 of colour of male is much deeper on both surfaces, chiefly because 

 there is very little of the white dusting, often so extensive and 

 conspicuous in the smaller race ; also the white spaces are less 

 extensive and they are notably so on hindwing ; on underside the 

 fore wings are distinctly black, instead of grey ; the hind wings 

 are of a -rich greenish-yellow, with white spaces sharply outlined 

 and bordered with brown ; in the nymotypical race these wings 

 are usually of a paler colour and the white spaces do not stand 

 out so boldly. A male from Turin and a female from the Simplon in 

 my collection come very near the Tyrolese ones, but they are not quite 

 so distinct. A series from Oberweidan, near Vienna, belongs, on the 

 contrary, to the nymotypical race, but it is a little larger and darker 

 than the Italian series mentioned above. Oberthur's fig. 475-6 in the 

 Et. Lipid. Comp., represents well Hiibner's race, and to the same 

 belongs his figure of form valesiaca, Riihl ; his nevadensis from Sierra 

 Nevada (fig. 474) seems to resemble speciosa in size and colour, but 

 being a female, one cannot be sure, and anyhow the latter lacks the 

 peculiar characteristics of the underside of hindwing and antenna? 

 observed by Rambur and quoted by Oberthiir. It will be interesting 

 to establish the distribution of two such distinct races of carthami in 

 Central Europe. 



Powellia sao, Hub., race gracilis, Vrty., first generation subgraciljs, 

 mini, and race sao, Hub., second generation parvula, mihi. — This 

 species seems to have been neglected by collectors ; I have never 

 received any specimens in exchange. The result is, till last year, I was 

 only acquainted with the race collected by myself and Querci in 

 Central Italy, which varies considerably individually and which 

 produces the remarkably minute second generation called by me 

 gracilis in the Ent. Reel, xxxi., p. 28, but which, on the other hand, 

 has quite the same aspect from the tops of the Apennines to the sea- 

 shore. When I saw a few specimens collected on the Coast Range of 

 Calabria and others from the Isarco V alley in S. Tyrol, I was so struck 

 by their appearance, new to me, that I at first took them to be orbifer, 

 Hb., which is recorded from the second of these regions and which is 

 found in Sicily. Closer inspection, however, soon convinced me I 

 was mistaken. I should mention that I have not been able to find 

 any confirmation of Kane's statement, quoted by Wheeler, that orbii'er 

 is found in S. Tyrol, either in nature or in the various local collections 

 I have examined. I next compared my specimens to Hiibner's figure 

 of sad, the habitat of which he gives as "Germany, in several regions," 

 and I found that it exactly answered to the races, so similar to each 

 other, of S. Tyrol and of Calabria. I thus found out that the race of 

 Central Italy, which I had called gracilis on account of its minute 

 second generation, differs from the nymotypical race also in the first 



