229, THE ENTOMOLOGIST. 
however, determined to investigate the point on the first oppor- 
tunity. This opportunity arrived this summer, when I observed 
L. orbitulus at several Swiss stations, and at Arolla met with it 
in sufficient numbers to enable me to follow the matter up. In 
none of these localities could I find any Androsace vitaliana; it 
was therefore evident that the food-plant of the butterfly in 
these localities must be some other species. The butterflies 
were not plentiful or the weather favourable enough to make the 
investigation an easy one, and I had to spend a whole day and 
portions of several others before I could satisfy myself as to the 
correct solution. In fact, I did not succeed in seeing one 
specimen lay an egg, but I found the females of orbitulus were 
always attracted to, and paid special attention to, one plant, 
and on one occasion, and perhaps on a second, an egg appeared 
about to be laid, but actually it was not done. The plant proved 
to be Soldanella alpina. Butterflies in captivity laid freely on 
this, and the young larve now hatched eat it readily. 
It is no doubt a fact well known to botanists, but to me it 
was quite a discovery to find that Soldanella alpina was an 
abundant plant over large tracts of the alpine pastures between 
6000 and 8000 ft., forming quite an appreciable portion of the 
herbage. The ordinary tourist, from whom botanically the 
average entomologist can hardly be differentiated, looks on 
Soldanella as only occurring at very high elevations (8000 to 
9000 ft.), and near to snow, and then rather sparingly. This is 
true, however, only of plants in flower in high summer, say 
July ; I must confess to having had some such ideas. 
The plents at lower elevations must flower very early. Not 
only the flowers, but also the fruits (if any), had disappeared in 
the haunts of orbitulus at the end of July. . 
Soldanella alpina is therefore the food-plant of L. orbitulus 
at many, if not a majority, of its habitats. On the Soldanella, 
as on the Androsace, the larva lives on the leaves and not on 
the flowers ; though assuming, as seems probable, that it hyber- 
nates half-grown and feeds up in the spring, it may at that time 
attack the young inflorescence. 
Unfortunately both my plants and larve have suffered by 
the unavoidable ill-usage of travel, and I much doubt if my 
material will enable me to carry the life-history further than I 
did last year. 
Soldanella, like Androsace, belongs to the Primulacee, so that 
it seems very probable that ZL. orbitulus may also feed on some 
other plants of that order, Androsaces or alpine Auriculas. 
Both plants have thick fleshy leaves, and the larva bores into 
these and scoops out the parenchyma through a small hole, 
much as a Coleophora would do, or as the flower-buds are treated 
by the young larve of argiolus, betica, and other flower-eating 
Lycenids. 
Betula, Reigate: August 16th, 1909. 
