124 A First Study of Natural Selection in Clausilia laminata 
which occurs round Hamburg, are common at Risborough, but absent from 
Gremsmiihlen. The small size of the Risborough wood, and the consequent 
absence of true forest conditions, permit the free entrance of H. variabilis. The 
only species of Clausilia which occur are G. laminata and G. nigricans. 
The species of Helix mentioned, and Glaiisilia nigricans, are found in the 
hedges or meadows which surround the little wood at Risborough, but Clausilia 
laminata occurs only in the beech-wood. A similar relation between Clausilia 
laminata and C. nigricans can be demonstrated on many of the English chalk 
downs. In the places where beech-trees occur, the two species are found together ; 
but C. laminata does not occur, or occurs but rarely, away from beech-trees, while 
G. nigricans is found in other places also. 
These two races of C. laminata show first that races with a sensibly identical 
spiral exist in localities so widely separated that no crossing between their 
ancestors can have occurred for an immense period of time, and so different that 
the environmental conditions common to both are comparatively few ; secondly 
that the existence of C. laminata is possible under conditions which permit the 
existence of G. nigricans; but that the latter species can live and multiply in 
places where the conditions necessary for the life of C. laminata do not occur. 
Now the differences between these two species are (1) a difference in size; (2) a 
difference in the pitch of the spiral ; and (3) other slight differences in the shape 
of the shell and of various organs. All these differences seem to us, at present, 
of little physiological importance, and yet the sum of these differences is 
demonstrably associated with an enormous difference in susceptibility to certain 
environmental differences : and since this is the case we are justified in accepting 
evidence which points to a correlation between variation in death-rate and 
variation in a character within the limits of one species, although we are at 
present quite unable to imagine the process by which this correlation is brought 
about. 
