This process, forming the bridge from mere cell division to sexual reproduction, is a hypothetical one; it has not, I believe, been actually observed in unicellular organisms, but it is exactly what we find to be taking place when we reach the stage of sexual reproduction among multicellulars. Multicellular organisms of more or less elaborate structure plainly cannot, without breaking up, fuse together like single cells. How, then, are they, as a species, to gain the advantages of the temporary union and interchange of elements which we have observed in the low unicellular organisms? Only in one way—by producing special cells for this purpose. These cells must represent the whole parent, they must be capable of shedding half their chromatin, and, when they have fused, must be capable of growing into a complete organism like the parent. When these specialized cells have been formed, the others, the somatic cells, will at the same time have been specialized for other functions, and will thus naturally lose the original capacity for interchanging chromatin with other cells, _i.e._ for conjugation. We see the significance, then, of Weismann’s remark, “germ cells made their appearance along with the multicellular body.”[43] They are an instance of that differentiation of structure and function which takes place in all highly organized life. We must note also that the benefits of conjugation which are realized _individually_ by the lowest unicellular forms are only realized _as a species_ by the multicellulars. A species must, then, be regarded as in some sense an organic whole, and not as a mere aggregate of individuals.