144 THE TREND OF THE RACE 
creasing mortality. If people live longer, there is naturally a 
larger number of them alive at any given time. If each family 
always produced the same number of children the relative num- 
ber of births per thousand would decrease as the number of 
people alive at any given time increased. Therefore, with the 
same marriage rate and the same degree of fecundity, a commu- 
nity with a decreasing mortality would show a decreasing birth 
rate, were we to measure birth rates, as is usually done, by the 
annual number of births per thousand inhabitants. 
Marriage rates estimated, as they commonly are, by the num- 
ber of marriages made annually per thousand of the population, 
would be changed by both the birth rate and the death rate. 
With a given number of marriages per annum, the rate per 
thousand of the population would decrease with an increased 
birth rate and increase with an increased death rate. In consid- 
ering the relation of marriage, birth and death rates it must be 
borne in mind that each of these affects the others as expressed by 
the method usually employed. 
Changes in the birth rate arising from variations in the rate 
and age of marriage and the death rate may be partly avoided 
by employing the so-called ‘“‘corrected births rates” in which 
allowance is made for changes in these factors according to the 
method employed by Newsholme and Stevenson or some similar 
mode of procedure. An index of birth rates for many purposes 
more satisfactory is afforded by the number of children born 
annually to every 1000 women of child-bearing age. What 
method of enumeration is the best depends on the particular use 
one wishes to make of the data. 
Statistics on the birth rate may also be vitiated to a certain 
degree by immigration and emigration. In the United States, 
not only foreign immigration, but the frequent emigration of our 
people from one state to another introduces a source of error into 
the statistics compiled by the several states. In addition, the 
vital statistics of our states suffer from other sources of inaccu- 
racy due to the way in which they are compiled. Data on births 
are faulty owing to incomplete birth registration. Only a few 
