According to our assumptions the behaviour changes parallel those of the physique, first as a consequence of a chance change in the latter, but very soon directing the further selectional mechanism into definite channels, because, according as behaviour has availed itself of the first rudimentary benefits, only further mutations in the same direction have any selective value. But as (let me say) the new organ develops, behaviour becomes more and more bound up with its mere possession. Behaviour and physique merge into one.
