It is no less true that statistics attest an excess of deaths over births. The information which they furnish has, however, been presented without distinction. Old and new creoles have been mixed together, as well as the latest emigrants, in a common estimate. Elements, which are fundamentally very different have thus been confounded. For a work of this kind to have any real value, it is absolutely necessary to divide the population into classes determined by the time of emigration, and to estimate the length of time itself by the number of generations. By proceeding in this manner, we shall undoubtedly establish in the mortality of groups striking differences, more or less analogous to those displayed by the generations of plants and animals transported into Africa or America. The statistics in question are still further vitiated by a fault, which is completely exposed by M. Walther in his work upon Guadeloupe. He, also, has drawn up tables of mortality; only, instead of taking the population _en masse_, he studied each district separately. Very significant differences then made their appearance. Considered as a whole the population of Guadeloupe offers an annual excess of 0·46 deaths over births, that is to say, nearly one-half per cent. In presence of these facts, the statisticians whose views I am attacking, would certainly have concluded that the European is not acclimatised in Guadeloupe, and have declared, that, after a certain time, which might easily be calculated, this colonial population would become extinct, if the voids were not incessantly filled by fresh immigrants. When, however, we examine the table of mortality taken by districts, we arrive at very different conclusions. These districts number thirty-one. Now, in fifteen the number of births is greater than that of deaths. In the little island of Marie-Galante this is the case in two districts out of three. Thus, the terrible calculations of the mean mortality are due entirely to the exaggeration of mortality in certain districts, while the European has become acclimatised in the others. The tables of mortality drawn up in Algeria by M. Boudin present analogous facts. Out of sixty-nine localities, fifty-five have shown, since 1857, an excess of births over deaths. The general result obtained by M. Walther may be thus explained. The French race is acclimatised in Guadeloupe in fifteen localities, but not in the remaining sixteen. Of these two statements, the first should be considered as definitely proved; the second requires confirmation, for a closer examination of the populations of the most unhealthy districts, and a study of them in classes, is still required. However this may be, every unprejudiced person will acknowledge that we can no longer question the fact of acclimatisation in _Guadeloupe_ as a whole. It should now only be a question of acclimatisation at _Basse terra_, at _Pointe-à-Pitre_, at _Pointe-Noire_, etc.