The interesting notice in your last number, of M. Coquand’s ‘Oysters of the Chalk,’ draws inferences unfavourable to the theory of development or evolution which scarcely seem warranted by the facts. It need not be ‘difficult to imagine the creature as existing under such conditions, that one species, while engaged in “the struggle for existence,” should starve out and extinguish another;’ for however widely we may find a fossil species dispersed, it is not probable that it occupied the whole of its territory at one and the same time, and in the limited area occupied immediately before its extinction, new varieties may have prevailed over and displaced the old by some slightly superior adaptation to the food-supply of the region. The extinction of any particular species may in some instances have been due to the extinction, or loss by other means, of its own appropriate food. Again, it is not necessary to suppose that the hinge, or the internal or external structure of the shell of an oyster, has been altered by what may be called the direct action of ‘natural selection,’ since by the well-established principle of ‘correlation’ the variation in one part of an organism is nearly or quite certain to produce variations in other parts. ‘If any such change did occur,’ it is argued, ‘it must have been _per saltum_, since with these mollusks, numerous as they are, there are no forms that can fairly be recognised as transitional.’ But this appeal to the evidence of facts is somewhat premature. The immense difference pointed out between the geological records of England and France in regard to these very oysters of the chalk, leaves it perfectly open for us to suppose that even the comparatively full French record is itself exceedingly imperfect, and that the transitional forms have either not been preserved, or remain yet to be discovered. Mr. Darwin gives reasons for believing that when variation once begins it continues with some vigour; hence, between two settled wide-spread species connected genealogically together we might expect a large number of transitional varieties, each represented by only a few individuals, so that the whole number of these transitional forms might well be lost to the genealogical record.