204 Roger Verity 
effaced as compared to australis and to ferrea (form detersa, 
Vrty., Bull. Soc. Ent. Ital., XLV, p. 226). Also the II generation 
differs in the same way, on upperside, from that of australis and 
varies, both individually and locally to about the same extent on 
both surfaces. In some localities, chiefly in the mountains, it is, 
to the naked eye (apart from the androconial scales), more or less 
identical to the I generation and I should name it postemiaustralis, 
from specimens of Dombresson, in the Jura. In other races the 
underside is like that of aestivalis; this I believe to be the usual 
form in the warmest localities of the Valais and all over southern 
France: infraestivalis, mihi (“type“ from Martigny). In the hottest 
and driest localities a form is produced parallel on underside to 
emilyllus and this Krulikowsky has named semilyllus. Some in- 
dividuals of the latter have on upperside a broad marginal band, 
as in mareinata, Rühl, but of a lighter, gray, tone: /atecana, mihi 
(“type“ from La Rognette in the Alpes Maritimes Departement). 
In the northern part of Central Europe (Northern France, Bel- 
gium and Northern Germany) there exists a race, which is like 
emiaustralis in size and look of upperside, but which has a more 
variegated underside, pointing to scofa in most individuals by the 
more pronounced white band and by the darker basal half of 
hindwing as compared to outer half: race centralis, mihi (typical 
series from Chantonnay in Vendee). Its II gen. postcentralis, mihi, 
differs from the I in most individuals of both sexes by the 
warmer chestnut tinge of underside and by its bolder markings, 
making it still more variegated. It is in this race (Bull. Soc. 
Ent. Belgique, 1914, p. 8) has discovered the difference between 
the androconial scales of the two generations, but he was baffled 
by the continuous emergence “during the whole of the good 
season“ and he was not able to detect them. He found spring 
"scales as late as June 20th and summer ones from the 26th. 
Nothing could confirm more perfectly the conclusion I have come 
to in Italy that there are only two generations, overlapping at 
the end of June and the beginning of July, but Ball fell short 
of realising this fact. He again found vernal scales at the end 
of September and remarks they probably were precocious autumnal 
individuals of the I gen. 
In the British Islands there is a race remarkabe by its very 
variegated underside. I have named its most extreme form scofd, 
in the Bull Soc. Ent. Italiana, XLI, p. 271, pl. I, fig. W 
(1911), (see Ent. Rec., 1916, p. 173), from an August series Ol 
the north coast of Scotland, but even in the South of England 
extreme individuals resemble it, whilst others are like centralis. 
