1906 Biometrika  
IT has been suggested that the approach to equality in male and female births is an illustration in some mysterious manner of Mendel's theory' of heredity. Observers have actually counted the number of males and females born in divers species with the conception that the approach to equality thus rendered manifest illustrates in some way Mendelian principles. I am not prepared to say it does not, because I have failed to grasp the manner in which those principles are applied to this case. If
more » ... to this case. If the demonstration depends, however, on the equality of the male and female births, their sensible inequality* in the case of man requires some further explanation; it is a case wherein environment or a priori, perhaps, race causes permanent and fairly constant deviations from equality. The aim of the present paper is to show that, as far as the writer can judge, there is no inheritance, Mendelian or other, of the sex-ratio. So far it confirms the results of Dr F. A. Woods stated in the previous paper, but the method of approaching the problem differs from his. No assumption is made as to the existence of a Gaussian distribution for the frequency, and the sex-ratio for the family of each individual is directly calculated and tabled. The paper further deals with the case of horse as well as man. There is no difficulty in extending the investigations to cattle and dogs from the herd and studbook returns, but the negative results provided by two such different species seem sufficient to demonstrate that the non-inheritance of sex is fairly widespread. The material is the following: (i) Data from a series of schedules on the size of families issued by Professor K. Pearson. Unfortunately this material proved less ample for this special purpose than we had anticipated. For although marriages must have existed at least 15 years in both generations for a schedule to be filled in, it happened in a very large number of the cases that the families in both generations did not provide the number (four) of children which seemed the least upon which a determination of sex-ratio could be made. Only 348 cases were taken from this source.
doi:10.1093/biomet/5.1-2.79 fatcat:y5our7j6kjfrrfmugb2hm5duia