LAW I.--_A medium type cannot exist by itself, except on the condition of being supported by the two creating types._ LAW II.--_When two types become united, two phenomena may arise: 1. Either one of them will absorb the other; or, 2. They may subsist simultaneously in the midst of a greater or less number of hybrids._ These two laws are only, in fact, the formula of the principles which Prichard[256] himself laid down long ago, and which are held also by the editor of the _Ethnological Journal_,[257] by Knox,[258] and by William Edwards.[259] By reason of these laws, we find that nowhere can a medium race either establish itself on the ruins of two creating races, or replace them, and live by itself with an independent existence, formed entirely of hybrids which propagate among themselves. In fact we have laid down a rule that conditions of development are very much restricted among hybrids, and that they can only go on decreasing in their descendants, if they are capable of producing any. The crossed race will only exist in the condition of being supported by the two creating types remaining in the midst of it. If the value of this law is only deduced from a negative fact,--that is to say, from the absence on the surface of the globe of any real hybrid race existing by itself on a certain extent of territory,--we shall find, as regards the second, a great number of _positive_ facts. When two types unite, we said, a double phenomenon may be observed; either of these two types will absorb the other, or they may subsist simultaneously, one near the other. The first case ought to be the most frequent; but it is the least appreciable, because it does not leave sensible traces. We must endeavour to discover in history the remains of a people who formerly existed, and who have since disappeared. Thus, the colony of Nubians, taken to the banks of the Phasis by Rámases, have left no trace of their sojourn among the inhabitants of the land. It is the same with the Greek colonies of the Mediterranean coasts.[260] The Normans have only left, on the coasts of Labrador, their engraved _stelæ_;[261] their race has not remained. The primitive Turkish and Asiatic type has likewise disappeared from Europe. This has been attributed to the introduction of Georgian women into the seraglios, and it is, perhaps, a reason only too readily accepted. It is, indeed, very natural that the repeated introduction of Georgian and Circassian women into the hárems should deprive the descendants of the conquerors of their original characteristics; but if this were the case, the Turks of our days would, from continued unions with the same race, have become real Georgians and Circassians themselves. It has not been so, however, because the harems are recruited in Europe as well as in Asia, and even then the fact would only be applicable to families of high position. The truth seems to be that the real Turkish blood has