has been inheritance of acquired characters or there has been no evolution."
Put shortly his arguments are as follows: In the individual increased functional activity in any structure is followed, generally speaking, by an increase in the size of that structure; this increase of functional activity and size puts a strain on, and is followed by a proportionate increase of activity and consequently of size in many structures co-ordinated with the first; these latter changes are the cause of changes in yet other structures; till, if the initial change be sufficiently important, the whole organism is co-adaptively modified. Now if acquired variations are transmissible, it is easy to understand how these acquired co-adaptive changes, dependent on functional activity, on use, accumulated during generations, may result in co-adaptive evolution of the most far-reaching kind in all the thousand co-ordinated parts of a complex and heterogeneous animal—e.g. mammal. But if they are not transmissible, by what means can we account for co-adaptive evolution? How can we explain the fact that a thousand structures are modified in the moose, the elephant, and the man, in obvious co-adaption to changes in the horns, the trunk, and the fore limbs respectively? For if we suppose that only inborn variations are transmissible, we must suppose that all the thousand structures varied favourably simultaneously and proportionately, and to suppose this is to suppose that which is incredible.
For instance, if during the evolution of the great horns of the moose an animal varied spontaneously in that he had larger horns than the ordinary, it is highly unlikely of the thousand structures, the evolution of which has been concomitant with and co-adaptive to the evolution of the horns, that all should have varied favourably, and if they did not the favourable variation
