On one hand we see the singular effects of castration, which we have already considered; on the other hand, an extraordinary constancy in the hereditary characters of the species. For more than three thousand six hundred years, which corresponds to about eight hundred generations, the Jews have been circumcised. Nevertheless, if a Jew ceases to circumcise his offspring the prepuce of his children grows as it did three thousand six hundred years ago, although, during the eight hundred generations in question, its absence from birth has prevented it reacting on the germinal cells of the individuals. If the engraphia of the external world could sensibly modify in a few generations the hereditary mneme of the species, it appears evident that the Jewish infants of the present day would be born without prepuce, or at least with an atrophied one. It is on such facts, which are innumerable in natural history, that _Weismann_ relies to repudiate absolutely the heredity of characters acquired by non-germinal organs and to attribute the development of organisms to blends and combinations due to conjugation, or crossing, as well as to natural selection, which he regards as all-powerful. _Darwin_ well recognized the difficulty in question, and being unable to explain the facts, had recourse to the hypothesis of _pangenesis_, that is of small particles detached from all parts of the body and transported by the blood to the germinal cells, to transmit to them, for example, the qualities acquired by the brain during life. This hypothesis was so improbable that _Darwin_ himself was forced to recognize it. Let us examine the facts. On the one hand a newly born Chinese transported and brought up in France will learn French, and will show no inclination to learn or understand Chinese. This well-established fact seems in favor of _Weismann_ and against the heredity of acquired characters. But, on the other hand, we cannot understand how the evolution of the brain and its functions takes place, without admitting that in one way or another the characters acquired by habits repeated during many generations gradually accumulate in the form of hereditary dispositions in the germinal protoplasm. It is certain that our brain has progressed since the time when our ancestors were similar to the gorilla, or even the cave man at the beginning of the quaternary age. How can this cerebral progression be explained only by selection which can only eliminate, and by crossings which by themselves can hardly raise the average? It is here that the intervention of an unknown power is necessary, something unexplained, the action of which has been lately recorded in the phenomena of mutations of _de Vries_.