288 THE TREND OF THE RACE 
fect is antecedent to their alcoholism. But because the bulk 
of the mentally defective became criminal or alcoholic it does not 
follow that every alcoholist is mentally defective, and will breed 
mentally defective children.” 
Another objection to the conclusions of Elderton and Pearson 
is that in no case was it certain whether or not the parents began 
their alcoholic habits before the birth of the offspring. It is a fair 
presumption, from what is known of the persistence of habits in 
human beings, that the parents who were alcoholic after their 
children became of school age were in most cases more or less 
alcoholic before their children were born. Of course the alcoholic 
habits of people are subject to much variation, and some parents 
may have used alcohol before their children were born and after- 
ward became sober, and in other parents the alcoholic history may 
have been just the reverse. To the extent that such changes 
occur, whatever correlations may exist between parental alco- 
holism and the characteristics of offspring would not be revealed 
by the statistical methods employed. The presence of fluctua- 
tions in the alcoholic habits of parents would naturally weaken 
the correlations that might exist between alcoholism of parents 
and peculiarities of their children. These correlations would be 
further weakened by the fact that the classes compared were not 
as sharply defined as would be desirable. The teetotalers were 
unfortunately very few in number and for statistical treatment 
they were usually grouped with the sober or those who drank but 
little. For the same reason the small group of those “‘suspected 
to drink”’ were combined with the drinkers. 
The investigation of Elderton and Pearson is of a type that it is 
desirable to see extended to further data. If the results do not 
justify a final verdict,—and the authors make no sweeping claims 
for the general applicability of their conclusions,—the fault lies in 
the inherent difficulty of the problem rather than in the imperfec- 
tions of the methods employed. The authors set about investi- 
gating a particular set of data bearing on a most important prob- 
lem, and they stated their precise findings and some conclusions 
that could and some that could not be drawn from their data. If 
