Studies in American History 
367 
Table I shows that, under the direct primary in comparison 
with a like period under the convention, the city’s share in 
the offices was reduced by 22 or 4.6 per cent, while the coun- 
try’s share was increased by the same amount. Five of the 
eight counties showed a loss for the cities and a gain for the 
country towns. A further analysis of the figures, which are 
given in Table II, results in the interesting discovery that 
under the direct primary there has been a remarkable corre- 
lation of distribution of offices according to population as be- 
tween city and country. The distribution of offices in every 
county except Kennebec has tended to approach more closely 
to the basis of the distribution of population. On this basis, 
the cities of the counties of Cumberland, Sagadahoc, and York, 
in the period prior to the direct primary, had been under- 
represented in county offices, while the cities in the other five 
counties had been over-represented. A study of Table II 
shows how nearly the under-representation in the one group 
and the over-representation in the other have been wiped out. 
Bangor furnishes a striking illustration. During the conven- 
tion era, 1901-1911, she held 32 to the country’s 28 county 
offices, while under the direct primary, 1913-1923, she has 
held 22 to the country’s 38. 
An analysis of Table III shows that the correlation between 
the distribution of offices and of population was maintained 
in the biennial period 1924-1925. In the eight counties the 
cities contained 48.6 per cent of the population and enjoyed 
47.5 per cent of the offices, lacking only 1.1 per cent of their 
just proportion according to population, while the country 
towns enjoyed only 1.1 per cent more than their just propor- 
tion. 
The tendency under the direct primary for representation 
as between city and country to approach the ratio of the 
distribution of population is further clearly indicated by 
Tables IV and V. During the last six biennial periods under 
the convention system, the cities won 14.3 less than their just 
proportion of representatives according to population, while 
under the direct primary they approached to within 3.2 of 
their just proportion on the basis of population. Table VI 
brings the statistics on distribution of senators down to 1925. 
The cities have had a slight advantage over the country towns, 
yet the deviation from the per cent indicated by the distribu- 
tion of population was only 4 per cent. 
