THE FALL OF THE BIRTH-RATE 19 



me to provide quite sufficient accuracy. Further, it 

 avoids the necessity for going on to the second stage 

 of calculating a standardised birth-rate, since k is it- 

 self a measure of fertility, and birth-rates standardised 

 on the basis of Tait's law must bear the same pro- 

 portions to each other as the corresponding coef- 

 ficients k. We may therefore give " Tait's coefficient " 

 only instead of the standardised birth-rates, and the 

 calculation of the coefficient is very simple. Putting 

 k equal to unity in the formula, find the values of f t 

 for the central years of the age-groups available ; u 

 these values of /,, calculate the births that would be 

 given by the married women. The ratio of the actual 

 (legitimate) to the calculated births is the value of k. 

 For some years quinquennial age-groups are avail- 

 able in the English Census, but for others the age- 

 groups are I$ but under 20 



20 25 



25 M 35 



35 ii 45 



To secure comparability the same age-groups must 

 be used throughout, and therefore this was the group- 

 ing used, though a quinquennial grouping would have 

 been better. The values of/,/ioo for calculating k are 

 0-325,0-275,0-20, o- 10. Thus for Connaught in 1911 : 



Age Married Calculated 



group women Coefficn births 



15-20 89 



20-25 i ,626 



25-35 l<* 



35-45 22,059 



5,743 

 Actual births 13,697 



*.!3.$9? 8 5 . 



5-743 



In the last columns of Table IV are given the 



