SEASONAL POLYMORPHISM. 7 
early families belong chiefly to elews with still quite a good percentage 
of aestivus; the fourth extraordinary generation comes still nearer 
phlaeas, the tails becoming very short and often as in the northern 
initia, which it resembles furthermore by its small size; there thus 
exists in Tuscany the entire scale of variation, completed by the 
existence of fuscata-caudata as an extreme variation in the race aestivus 
of the Isle of EKlba: in my series of 34 males from this locality, five 
are perfect specimens of it; the rest vary much less than on the conti- 
nent, and, more or less, all belong to the form aestivus except early 
specimens of May. Zeller first described’ the very dark summer brood 
of southern Europe from series of ‘‘the mountains above Messina,” 
and Tutt rightly points out (page 371) that this name, although un- 
fortunate for a race, must be used for the southern one. I now add 
the suggestion that it should be restricted to races similar to the 
Sicilian one, in which aestivus largely predominates and fuscata-caudata 
makes its first appearance as an extreme individual form, whereas race 
nigrioreleus never produces it. Finally, when fuscata-caudata is abun- 
dant or predominates the race should take this name: it is more likely 
to occur in the eastern Mediterranean, as suggested by Tutt, but two 
out of three specimens collected by me during a short excursion in 
a little marsh near Tempio, in Sardinia, belong to this form; I was 
deterred from enquiring further into the matter by the presence of 
several bulls nervously inspecting me and my net. 
I need scarcely say that the geographical areas mentioned above 
are set down on broad lines, for, locally, one may.find in each, races 
which come near the races of other regions. In the hottest Alpine 
valleys the form aestivis is frequent, but 1t occurs mixed with the 
extreme opposite variations ; thus, the medium is brought back to the 
elews grade in these very variable races, such as are those of many species 
in the Alps: varieleus. On the contrary, at the highest altitudes of 
phlaeas in the Apennines, melanism is as limited as in Central Kurope, 
and the same names elews, and even initia-caudata, are well suited to 
it. As a general rule, at equal temperatures, dampness increases 
melanism, so that it often increases markedly in the Alps and con- 
stantly in the Islands and in marshy localities. 
The table given above shows that in this species seasonal and geo- 
graphical variations are identical. In the production of race-charac- 
ters in the Lepidoptera, phenomena of two orders generally combine: 
heredity and surroundings. It is clear that in this species the latter 
alone produce them by their action during the development of each 
individual ; even here, however, a slight hereditary difference seems to 
have been produced, because Weismann [Fnt., xxix., p. 75] claims to 
have proved by experimental breeding that extreme northern races 
respond more to cold, and southern races to heat. The races hypo- 
phlaeas and chinensis show, besides, some characters which seem 
hereditary; they might be called ‘‘ paynocunertic” races, as compared 
to the ‘‘ ontocEnetic’’ European ones described above, which alto- 
gether consist of only one phylogenetic race, equivalent to hypophlacas 
or to chinensis. A high degree of ontogenetic variability is perhaps a 
prelude to phylogenetic scission or differentiation ; it evidently is an 
index of sensitiveness to surroundings, and one would naturally take 
it to mean a state of instability. Most Lepidoptera produce individuals 
or entire generations which resemble other kindred races, but a periodical 
