ON WOMEN S LABOUR. 



363 



Dr. Reid also supplied the more detailed information leading to the 

 following table and calculations by Mr. A. L. Bowley : — 



These details make it possible to calculate a cori'elation coefficient 

 between occupation and mortality. 



Taking the columns «i and d only, and counting the towns as of equal 



importance, the formula — t — [where n is the number of towns, («, S,) 



9i (Ti ff2 ' vr 1 1/ 



(/i2 82) . . . the deviations in columns m and d from their average, and 

 ffi, (Tg, the standard deviations of these columns,] gives -47, witli probable 

 efvov •! ; this gives evidence of some con-elation, but, as the coefficient 

 is only five times its probable error, the evidence is not very conclusive. 

 If we now weight each deviation with the population (column j)), the 



fornmla — "'V gives -.57, with probable error "07 (the standard devia- 



tions being calculated afresh). 



We may say, then, that the excess or defect of proportionate occupa- 

 tion of married women is shown by these statistics to have a relation to 



' Leek is in advance of the other towns in sanitary matters ^nd is in a JieaJthy 

 position. 



