It is known that while the secondary sex-ratio of a species is fairly constant and not far from equality it is distinctly variable. It varies with the species.
It is possible by selection to obtain strains differing widely in respect of their secondary sex-ratios. King (1918) has found that in the case of the albino rat it was possible, starting with two pairs of rats from the same litter, to found two strains, one of which produced a high proportion of males, the other a prepon derance of females. The progeny of one pair (pair A) were bred brother to sister without selection for six generations in order to build up a homozygous and uniform race. After this, selection was practised, the brothers and sisters being chosen from litters which showed a preponderance of males. In line B the selection after the sixth generation was made from litters showing a prepon derance of females. After fifteen generations of such inbreeding and selection, the sex-ratio at birth in line A was 125:10O, in line B 83 :100. The habitual production of an unusual sex-ratio can be the expression of the hereditary constitution of a stock. It is possible to breed for a preponderance of one sex. If this process can be developed successfully in the case of domestic animals, breeding for sex preponderance will assume economic importance. It varies with the seasons of the year. It is not improbable that the breeding season is one in which the general physiological con dition of the individual is above the average, and it is of interest therefore to compare the secondary sex-ratio following conception during the breeding season with that following conception at other times. The human birth-rate actually shows a slight variation in spite of the fact that nearly all traces of the primitive breeding season have been obliterated by social habits. It is found that the sex-ratio is low for births resulting from conceptions at the seasons of greatest fertility and high at the time of lowest birth rate. In the case of the fowl there is evidence to show that there is a swing in the secondary sex-ratio with the chronological order of the egg. Jull (1923), using a sex-linked cross in order to preclude errors, recorded the sex of the chickens hatched from eggs of 45 hens during their first year of production. The observations were
repeated for three years and the secondary sex-ratio was found to be 48.41 expressed as a percentage. Analysis of the figures gave the following table: The secondary sex-ratio is profoundly disturbed as a result of interspecific and intervarietal crosses. Haldane (1922) has pointed out that in any such cross the sex that is absent, rare, or sterile is the digametic sex.
The effects of hybridisation upon the sex-ratio is a subject of considerable anthropological interest. Three types of hybrid isation are possible: (I) between white races, (2) between coloured races, and (3) between white and coloured races. As a result of an investigation of the effect of hybridisation, Little (192o) concluded that a higher preponderance of males resulted from hybrid white matings than from pure white matings, and that hybrid coloured matings gave a higher preponderance of females than did pure coloured matings. This is in agreement with Lewis' results in the Argentine (1906). Powers (1877) states that, in the case of black and white hybrids, there is a large excess of girls among half-breeds in California and Kohl (1859) notes that in the northern parts of the United States females preponderate in the progeny of French men with Indian women. Starkweather (1883) found in the mulatto a 12 to 15 per cent. excess of females, while in the whole population males were in excess. An excess of females is reported by Gdrtz and Waitz among the offspring of Dutch men and Malay women in Java (1859). Jastrzebski states that for the years 1910-1915 the ratios of males to ion females in New York were : whites, 104.0 ; negroes, 99.9; and mulattoes, 97.9. Bugnion (1910) states that in a hybrid race formed by settlers in a colony of negroes there was a marked preponderance of females. It seems therefore possible to conclude (I) that crosses between white races produce an ex cess of males over pure white matings; (2) that hybridisation between coloured races produces an excess of females above pure coloured matings; and (3) that hybrids of white and coloured races show an excess of females above the pure matings of either race.