Selection of Helix nemoralis by the Song-Thrush. 527 
taken from A, none fram B, three from C, five from D, and 
five from EK, although the two latter groups were not put 
out until four days after the others. Unfortunately it was 
not possible to make longer observations, but, as far as they 
go, they suggest: (a) that the thrushes, so far from choosing 
the lightly banded shells, actually preferred the four-banded 
variety, although this was not the most numerous form 
provided ; (4) that thrushes may break snails which for some 
reason they do not eat; (c) that there is a tendency for 
snails to be taken from the more exposed places. If the 
distribution of H. nemoralis at Quy Fen obtains elsewhere, 
the latter factor alone might account for the large proportion 
of unbanded shells at the “ anvils” recorded by Mr. Trueman. 
A second series of observations was carried out near 
Cambridge round an artificial piece of water. The place, 
which is of about 4 acres area and surrounded by fields, is 
lush-grass thickly interspersed with cypress, privet, and 
bramble-bushes. In the season in question it was inhabited 
by one pair of thrushes, who reared two broods there. 
Collections were made at intervals , with the results set out 
in Table II. On July 31st anvil I. only was taken. On 
August 13th it was emptied again, and II., III., and IV. 
were found. Therefore the numbers for anvil I. on this date 
represent a proportion of the thrushes’ takings for a fort- 
night. The large proportion of Helicella cantiana found 
broken may possibly be accounted for by the dry weather 
that prevailed at the time. In August attempts were made 
to form a control collection, but owing to the drought they 
were unsuccessful, and further opportunity did not arise 
until October 15th-21st, when, after long search, living 
(mostly immature) specimens were found, as recorded on 
Table IIT. 
While the control collection is thus so small, judgment 
must be suspended ; but so far there seems no ground for 
supposing that the banded shells were taken less fr equently 
than the unbanded, nor that, when taken, they were not 
eaten. After my observations on the tethered snails, I was 
prepared to find that the heavily banded shells were dis- 
tasteful to the birds ; but the later observations dispose of 
this idea, and as these anvils were made in the summer, when 
insect food was plentiful, it could hardly be that hunger 
drove the birds to eat unpalatable food. 
I should like here to express my indebtedness to Mr. H. H. 
Brindley, M.A., of St. John’s College, for his assistance in 
forming these collections and for much helpful criticism 
and advice. 
