ANALYSIS AND INTERPRETATION OF DATA 83 



purpose of throwing some light on this subject was 

 striking in its results, though the conclusions, based on 

 two hundred and twenty-five cases, were not as certain as 

 might be desired (Table XXV). In this table the men 

 of letters born into families containing a given number 

 of children were classified according to their rank in the 

 group of their brothers and sisters. If their birth-rank 

 were purely a matter of chance, in families of a given 

 size there would be equal numbers of literary children in 

 each rank, from first to last born. Thus, for example, 

 in families of five children the normal probability is that 

 each rank would contain twenty per cent of the total 

 number. Actually it was found that of the literati born 

 in such families nine were first-born, four second-born, 

 four third-born, six fourth-born and eight fifth-born, or 

 twenty-nine, thirteen, nineteen and twenty-six per cent 

 respectively. Such a process of analysis was carried out 

 for eleven sizes of family, from two up to " twelve and 

 over ". In only three of the eleven classes did the num- 

 ber of first-born fall below the number to be expected. 

 In those three cases the number of families concerned 

 was small, and chance fluctuation might well account for 

 the result. Even so, the discrepancy betweeen the act- 

 ual results and the normal probability was not large. 

 Likewise in only three of the eleven classes did the pro- 

 portion of last born fall below the number to be expected. 

 On the other hand, in six of the eleven cases the number 

 of second born fell below the probability; and, in eight 

 of nine cases considered, the class " third and later born, 

 not including last-born/' likewise fell below the normal 

 probability. 1 



1 In addition to the literati who could be accurately ranked, there 

 were records of literati in families of unknown size, of whom eight 

 were first-born, seven second-born, and eleven born in all the other 

 ranks combined. In twenty-one cases the author was an only child. 



