[ 207 ] 
VII. On some Variations of Cardium edule apparently Correlated to the Conditions 
of Life. 
By William Bateson, M.A., Felloiu of St. John’s College, Cambridge, and 
Balfour Student in the University. 
Communicated by Adam Sedgwick, M.A., F.R.S. 
Received May 13,—Read June 6, 1889. 
[Plate 26.] 
Introduction. 
The following paper forms part of an investigation of tire relation between the 
variations of animals and the conditions under which they live. It appears to me 
necessary that any investigation of this problem should be begun by the examination 
of cases in which difference in environment is known to exist, and that variations 
should then be sought for among the forms of life subjected to these conditions. If 
by this examination any variations can be shown to occur regularly with the change 
of conditions, or in any way in proportion to their intensity, it is so far evidence that 
there is a relation of cause and effect between them. 
By thus first approaching the question from the point of view of the conditions, 
many difficulties are obviated which occur in any attempt which begins by ascertaining 
the variations in the animal, in the hope of afterwards finding an environmental 
change to which they may be traced. Such attemjits to trace back variations to 
some eiivh’onmental cause have often been made, and have, in general, been unsuc¬ 
cessful. In the case of species which have varied in isolated situations not apparently 
differing from each other, the failure to find points of environmental difference has 
been held to be evidence that the variations in question did not arise from such causes 
at all. This appears likely, and is probably true of the variations in question ; but it 
must be borne in mind that the fact that no palpable difference can be found between 
the conditions in the several localities is no proof that they do not exist. While 
these difierences in condition are usually evasive and hard to detect, it is best to begin 
to investigate their relation to variations in animals by selecting cases in which the 
change in conditions is unequivocal, and proceed from this starting point to seek for 
correlated variation in the forms of life subjected to them. 
It appears that a particularly favourable opportunity for investigating this question 
MDCCCLXXXIX.—B. 2 Q 31.12.80 
