agents, but that the evolution of multicellular organisms has resulted solely from the action of Natural Selection—from the natural selection of inborn variations alone—or, as Mr. Spencer holds, from the selection of inborn plus acquired variations.
If we hold, as I think we must, that the multicellular organism is really a compound being, a being compounded of unicellular organisms, then, as already explained, we must suppose that acquired variations are not transmissible, unless it can be shown that the reproductive cells receive from the somatic cells elements which so alter their constitutions as not only to cause them to proliferate into organisms different from those which would have otherwise arisen, but also into organisms which have inborn in them variations which were acquired in the cell-community of which the germ cells were members. This is the next question discussed by Mr. Spencer, but strangely enough he turns away from the actual point at issue, the question as to whether the elements (gemmules, physiological units, or what not) which are supposed to cause the change in the constitution of the germ cell, have any real existence; he assumes that they do exist, and appears to think that all he needs to prove is that, being existent, they are able to enter the germ cells, and this he does by pointing out that the microbes of disease as well as food molecules are able to enter them. He proceeds—
