My dear George

I have never supposed otherwise than that the gemmules breed abundantly all over the body, though I look upon them merely as local parasites, so to speak, that live, multiply & die in great multitudes in the places where they are lodged, though occasionally some of them may be detached and drifted along with the circulation, and so find their way to the sexual elements,— as was explained in the second part of the paper.¹

It is by the abundance of all sorts of them, in every part of the body, that I accounted in my paper for the reproduction of mutilated parts and other specified phenomena, adding “It w^d. much transcend my limits if I were to enter into these and kindred questions, but it is not necessary to do so, for it is sufficient to refer to M^r. Darwin’s work, where they are most fully & carefully discussed, and to consider while reading it, whether the theory I have proposed, could not as I think it might be substituted with advantage for Pangenesis”

(I have not the Contemporary Review by me & cannot give the page of the extracts— My copy is merely a revise, paged from 1 onwards. It is in the 12^th. page of the revise.)²

In this passage, I meant to include propagation by buds. You will see in the preceeding page, an allusion to the way in which the scattered alien germs “thrive and multiply”.³

Now for the application of all this: Wherever in a plant developed out of a bud or seedling, (no matter which, for the ‘stirp’⁴ is similar in both cases) the alien, localised germs happen to be congregated in sufficient number and varieties to form material for a fresh stirp, there will be a tendency to produce a bud. Structural conditions, such as those found at the parts where buds usually shoot, must of course be helpful in forwarding this tendency.

The advantage of my theory appears to be this: By Pangenesis, we should expect all animals however highly organised, to throw out buds. By my theory, I argue that where the animals are complex, the variety of germs concerned in the making of them, must be proportionately great, and consequently the probability of a complete set of them being anywhere in existence, in the same immediate neighbourhood, is diminished. Hence, the lower the organization, the more freely does it bud and the higher ones do not bud, which is in accordance with fact.

The budding even of the highest animals in the embryonic stage, is intelligible by the joint action of 3 causes special to that period.⁵ (1) The differentiation is less complete, and germs destined to be separated are then together. (2)The embryo being small, the alien germs in separate structures are nearer, than they become afterwards, (3) The tissues are softer and afford less obstacle to the approach and aggregation of the germs under their mutual affinities.

I hope I have answered fully enough, & much regret that I misunderstood the question, as put in your father’s letter,⁶ and have given you both, unnecessary trouble

I am eager to receive criticisms— even adverse ones.

Ever Y^rs. | Francis Galton.

About your father’s plants & the statistics of growth.:—⁷ In cases where not only the one biggest of each sort, but the two or three biggest were measured, the uncertainty of the relative values of the moduli of variability of the two sorts would be materially diminished.