By the way the notion of race never was based on the premise of a purity without any mixing.

For example John R. Baker in his 1974 book "Race", in chapter "The meaning of race", wrote:

"(...) The reader is aware that in following a widely distributed animal over its range, one often notices alterations in its appearance. The changes are particularly evident wherever a partial geographical barrier intervenes, such as a range of mountains, a desert, or a wide river. The populations on the two sides of the barrier are not, however, entirely distinct. Intermediates are found, and there is often direct evidence of interbreeding, though not on a sufficient scale to make the two populations indistinguishable. If the two populations are so distinct that one can generally tell from which region a specimen was obtained, it is usual to give separate names to the two races. If every specimen could be identified with certainty as belonging to one population or the other, it would be evident that no gene-flow occurred between the two, and they would therefore be regarded as different species in the genetical sense of the word, however small the differences might be. It is the fact that intermediates do occur that defines the race. (...)"

Then he added:

"(...) For practical purposes it may be found convenient to make an arbitrary decision as to the proportion of intermediates that are allowable, if different races are to be recognized. One may argue that a population ‘A’ is distinguishable from a population 'B’ if x% of the individuals constituting population ‘A’ can be recognized as not belonging to population ‘B'. It will be understood that the correct value to be assigned to x cannot be discovered by objective means; nevertheless, if a high figure (perhaps 75%) is agreed upon by taxonomists, one can scarcely doubt that there is a distinction worthy of recognition as subspecific or ‘racial’. Very commonly, however, the differences observed are so evident, and x is clearly so high, that no statistical investigation is necessary to convince other taxonomists that races should be distinguished. (...)"