At this point Westermarck finds it necessary to consider the problem of the "causes which determine the sex of the offspring." The view that sex is influenced either by the relative or by the absolute age of the parents is untenable;[418] nor can the theory be accepted that "polygyny leads to the birth of a greater proportion of female infants."[419] The theory of Düsing, however, must be regarded as the most probable explanation which has yet been advanced.[420] According to him, "the characters of animals and plants which influence the formation of sex are due to natural selection. In every species the proportion between the sexes has a tendency to keep constant, but the organisms are so well adapted to the conditions of life that, under anomalous circumstances, they produce more individuals of that sex of which there is the greatest need. When nourishment is abundant, strengthened reproduction is an advantage to the species, whereas the reverse is the case when nourishment is scarce. Hence--the power of multiplication depending chiefly upon the number of females--organisms, when unusually well nourished, produce comparatively more female offspring; in the opposite case, more male."[421] The observations of Ploss[422] and others[423] appear to sustain Düsing's hypothesis. Wherever nourishment is scarce there seems to be a surplus of male births. Such is the case in highlands as compared with lowlands; among the poor as compared with the rich; in sterile regions as compared with those that are more fertile. Furthermore, Düsing has suggested a second cause due also to natural selection, which influences the numerical proportion of the sexes born; and his conclusion is confirmed by the researches of Westermarck. Mixture of race among animals and plants appears to cause a surplus of female births;[424] while, on the contrary, incestuous unions, being injurious to the species, "have a tendency to produce an excess of male offspring."[425] So, among half-breeds, the number of girls usually predominates;[426] while among in-and-in bred plants, animals, or men the reverse is the case. Hence it seems probable "that the degree of differentiation in the sexual elements of the parents exercises some influence upon the sex of the offspring, so that, when the differentiation is unusually great, the births are in favour of females; when it is unusually small, in favour of males."[427]