HEEEDITAEY DEAFNESS. 
A DISCUSSION OF THE DATA COLLECTED BY 
DR E. A. FAY IN AMERICA. 
By EDGAR SCHUSTER, M.A., Francis Galton Research Fellow in 
National Eugenics in the University of London. 
The present paper is an attempt* to apply the newer methods of statistics, 
introduced by Mr Francis Galton and developed by Prof. Karl Pearson, to a 
valuable collection of hereditary data that have hitherto only been discussed by 
the older ones. 
The object is to obtain results which admit of direct comparison with others 
derived from quite other collections. These results take the form of " correlation 
coefficients," which give a numerical measure of the intensity of heredity. 
The best material for this purpose would have been a random sample of the 
population, of which the following particulars were known, namely the facts which 
relate to the presence or absence of deafness, and of its nature when present, 
among their parents, children, brothers and sisters, etc.; and not only this but also 
the total number of their relatives in those degrees in which the number is 
variable. 
rfeafness of the degree with which we are dealing is fortunately of such rarity 
that a sample of this kind, in order to be useful, would have to contain at least a 
million persons, so that practically speaking it is outside the range of possible 
attainment for a private individual. 
In default of this one is reduced to processes involving many possibilities 
of error, as one has to supply by estimate the deficiencies in such information 
as is accessible, in order to obtain artificial samples for making correlation tables. 
It might be objected that the facts taken have already been stated in a form 
that anyone can understand, and are now presented in a way that is only intelli- 
* I have to thank Prof. Pearson for giving me very much help in devising the special methods 
employed in this paper ; my thanks are also due to Mr Galton for reading the manuscript and 
suggesting many improvements in its form and style. 
Biometrika iv 
59 
