SERIES CORRELATED WITH GROSS EARNINGS 213 
each of the series. It follows from the above that if ratios of 
loans and discounts to earning assets are high when ratios of gross 
earnings are high, ratios of investments to earning assets are low, 
and vice versa, and that if ratios of demand deposits to total 
deposits are low when ratios of gross earnings are high, ratios of 
time deposits to total deposits are high, and vice versa. However, 
if ratios of demand deposits to earning assets are low when gross 
earnings ratios are high, it does not follow that ratios of time 
deposits to earning assets are high. Indeed, they tend to be 
low when gross earnings ratios are either high or low. 
Table 131 makes it possible also to determine, for each dis- 
persion-group of gross earnings ratios, the associated net disper- 
sion in the other series of data. Not only, for instance, are ratios 
of loans and discounts high or low, respectively, when gross earn- 
ings ratios are high or low, but the higher or lower the second, 
the higher or lower the first. These relations also obtain, in gen- 
eral, for ratios of total expense, salaries and wages, taxes, “other 
expense,” and net earnings, each series being expressed in terms 
of earning assets. 
In general, therefore, and with respect to ratios of total expense 
and of net earnings, for example, wherever and whenever— 
“wherever” relating to separate districts and “whenever,” to 
separate years—during the years 1919 to 1925, districts had 
gross earnings which were high or low, they also had total expense 
and net earnings which were high or low, both positions and 
amounts, for the series paired with gross earnings, being de- 
termined algebraically. Moreover, the higher or lower the gross 
earnings ratios, the higher or lower, respectively, the total expense 
and net earnings ratios. These relations, of course, do not obtain 
in every district every year. It is the prevailing tendency which 
is certain, and to which the summaries in Table 131 point. The 
details from which the net average percentages of ratios of total 
expense and of net earnings are drawn are given in Table 132. 
Both series of data vary directly with gross earnings, the rela- 
tion being more clearly marked in the case of total expense. The 
frequencies apply to the twelve districts for each of the seven 
years 1919-1925. But what is true for the total, as thus built 
up, tends to be true for the individual years for a single district, 
and for a single year for individual districts. Both facts are evi- 
dent from the scatter of the figures in bold type and of the letters