232 THE ALCOHOL PROBLEM 
in amount owing to the general shortage, was con- 
sumed in moderate quantities spread evenly over 
the week. 
Other evidence was obtained by comparing the 
frequency with which accidents were incurred by the 
night shift. These workers came on at 7.30 p.m., and 
if they took alcohol at all they were bound to consume 
it shortly before starting work, as they had no chance 
130 (r— = 
EE —— 
20 
] MEN 1915 
[ 
0 
< 
Q 
& 100 
8 
v) 
2 90 
ad 
2 
J 80 
~ WOMEN. 1915 
pico | 
Lm 
‘00 
-_—1tl— 
MEN 1916~17 pe en 
Serf | 
L 
M T WwW Th F£ § MT W Th FS 
17024. Dany verahon in accldents during 1246r and 1047: day. 
Fic. 24.—DAILY VARIATION IN ACCIDENTS DURING 12-HOUR AND 
10-Hour Dav. 
50 
of getting it during the course of the night, or next 
morning. Hence any alcohol effect, if it existed at 
all, would be most marked during the first spell of 
work, and least marked during the third or last spell. 
We saw in Chapter IV. that owing chiefly to increasing 
restrictions on the sale of alcohol during the war the 
convictions for drunkenness rapidly fell, and in the 
years 1916 and 1917 they were respectively 42 and