i8 THE FALL OF THE BIRTH-RATE 



Sir Arthur Newsholme) and Dr Stevenson in 1905 

 (refs. 4, 5*) and slightly later by myself (ref. 6). Drs 

 Newsholme and Stevenson used for their standard- 

 ising fertility-rates figures obtained for Sweden in 

 1891, and for the fixed index birth-rate the index 

 birth-rate for England and Wales in 1901. 



I used fertility data for Glasgow and Edinburgh in 

 1855. This was the first year of civil registration in 

 Scotland, and the age of the wife was then required 

 to be returned at the time of the registration of the 

 birth. This requirement was afterwards dropped, so 

 that we have no later data. Dr Matthews Duncan in 

 1866 had the data reduced for the two towns men- 

 tioned (ref. i) : the data were never officially compiled 

 and were not reduced as a whole till 1906 (Lewis, 

 ref. 3). Dr Duncan handed over to Professor Tait the 

 reduced data, and Professor Tait contributed some 

 very interesting mathematical chapters to the book. 

 He pointed out that an extremely simple law seemed 

 to express, within the limits of error, the fertility of 

 these married women : 



;< The percentage of wives of any one age who are 

 mothers within one year varies directly as the differ- 

 ence between that age and 50." 



In symbolic form, if f t is the percentage of wives 

 of age t who are mothers within one year 



/, = k ( 5 o - *). 



For the Edinburgh and Glasgow women Professor 

 Tait found k, which may evidently be treated as a 

 kind of coefficient of fertility, to be 1-5. Clearly this 

 law is no more than a rough approximation to the 

 facts (cf . the discussion in the paper cited) but it is 

 an exceedingly convenient one to assume for working 

 purposes on account of its simplicity, and it seems to 



* See list of references at the end of the paper. 



