died out completely in the course of several weeks. Those that had conjugated showed great variation (as usual); some died very quickly; others multiplied very slowly and finally died out; others multiplied more vigorously than any of the non-conjugants. At the end of six weeks, all those that had not conjugated were dead, while certain lines of the others had multiplied and were numerous. The difference between the two sets was in fact very striking. But it is important not to misunderstand the nature of this difference. The lot that had conjugated showed great variation, and many of the lines were not stronger than the non-conjugants, dying out fully as quickly. But a few were stronger, and these multiplied and replaced the rest. Thus after some weeks, all the survivors had come from hut three or four among those that had conjugated.
But even in these the depressed condition had not been completely overcome; they were still notably less vigorous than the strain which had been kept throughout under more natural conditions and had conjugated frequently.
Thus what had happened was this: Conjugation had produced much variation; some few of the variants had been more vigorous and had lived, while the rest died.
This result when first reached was unexpected and difficult to interpret. It seems of such importance that one felt it necessary to try it again. I shall not describe to you the long and wearisome process of providing anew the necessary conditions and repeating the experiment. It will suffice to say that the experiment was repeated and gave the same results as before.
Thus I believe that we are in position to make certain positive statements as to the effect of conjugation. Conjugation does not rejuvenate in any simple, direct way. What it does is to produce variation; to produce a great number of different combinations, having different properties. Some of these are more vigorous, others less vigorous. The latter die, the former survive. This happens equally, whether the animals which conjugate are at the beginning vigorous or weak. If they are vigorous, then one of the most striking effects of conjugation is to produce some lines that are less vigorous than the original ones, so that they die out. If the animals which enter conjugation are weak, then one of the most striking effects of conjugation is to produce certain combinations that are more vigorous than the original ones, so that they survive, while those that did not conjugate die out. In a short time the entire race is replaced by the descendants of a few of those that conjugated.
Now, the relation of all this to certain things that are known in higher organisms seems fairly clear. In higher animals likewise the result of intercrossing is to produce variation. We don't call it variation nowadays, because we know something more about it; we call it