154 Prof. K. Pearson. 



were not very copious, and in default of a method of dealing 

 quantitatively with characters not capable of exact scaling, it was 

 not possible to deduce absolutely conclusive results. Still Mr. Galton 

 brought good evidence to show that temper and artistic instinct were 

 inherited characters. On November 19, 1899, a paper was read to the 

 Royal Society showing how the inheritance of characters not capable 

 of exact quantitative measurement might be deduced. In that paper 

 I dealt with Mr. Galton's statistics, and showed that the fraternal 

 correlation in the matter of temper was O3167, and the parental corre- 

 lation in the matter of artistic instinct was O g 4039. These numbers are 

 somewhat low and not altogether satisfactory. I purpose in this 

 preliminary notice to give only a few results from some very elabo- 

 rate observations which have been made in the course of the last few 

 years. 



(2.) The material was collected in two separate ways. In the first 

 series the Family Measurement Series only physical characters 

 were observed. This series was started six years ago, and upwards 

 of 1100 families, father, mother, and not more than two sons and two 

 daughters, were measured. The series was closed two years ago, and 

 last year Dr. Alice Lee completed the reduction of this very large mass 

 of material. In its reduced form seventy-eight correlation tables have 

 been formed, giving as many correlation coefficients bearing on direct 

 or cross heredity. This is probably the most extensive series of inheri- 

 tance coefficients each based, as a rule, on upwards of 1000 pairs 

 which has yet been obtained. 



My second series will be still more extensive ; but it relates only to 

 collateral fraternal heredity. It aims at observing a wide range of 

 both physical and mental characters in pairs of school children. I 

 have received most kindly aid from a great number of masters and 

 mistresses in public schools, high schools, secondary and primary 

 schools of all classes. This will be very fully acknowledged in the final 

 publication of the results. But although the work has been in progress 

 for three years, we have still only material enough to draw conclu- 

 sions in the case of pairs of brothers, of whom more than 1000 cases 

 have been observed. 



The work has been carried on with the assistance from the Govern- 

 ment Grant of a sum appropriated to this purpose in 1898. Without 

 this aid it would not have been possible for me to purchase the neces- 

 sary head-spanners or to circulate them among the schools. 



(3.) Only three of the physical measurements of this extensive 

 series have yet been reduced, and the sister-sister and sister-brother 

 observations will have to be carried on for another year or two before 

 they are sufficiently numerous. The whole material will then require 

 two or three years for tabulation and calculation. But as the problem 

 of the inheritance of the mental characters and their correlation 



