306 THE TREND OF THE RACE 
families of a given size the first and second born show as a rule a 
preponderating amount of albinism, criminality and tuberculosis. 
Mongolian idiocy was found to characterize in a rather striking 
manner the last born of the family. 
When we investigate the incidence of any quality in regard to 
order of birth in individual families we are not entirely free from 
statistical pitfalls, if we start with material segregated in institu- 
tions. If we take individuals of a certain age, say 20, which are 
confined in a sanitorium, then if the numbers of families are 
increasing in the population at large the individual will be more 
apt to be the eldest of a recent family than the younger member 
of an old family. This possible source of error was pointed out 
by Mr. Cobb who says: 
“Tt has hitherto been assumed that if a person of given age is 
selected at random from amongst fraternities of a given size then 
all positions in that fraternity are equally likely. But this is 
not the case. If the number of births has been increasing he is 
more likely to be one of the older members of his fraternity, and 
if the number hasbeen decreasing he is more likely to be a younger 
member. For while the number of births is increasing there are 
more children born every year who belong to the first half of their 
fraternities than who belong to the second half.” 
In most countries there are more births per annum than 
previously and a steady increase in the number of families. But 
granting that this would give us an apparent increase of the first 
born of any particular age there is a compensating tendency 
brought about by the declining birth rate. Along with an in- 
creasing number of people there has been a reduction of the 
percentage of the later born owing to the increasing restriction of 
the size of the family. Consider a random group of 20 year old 
individuals from families of twelve members. Will not this be 
more apt to represent the last members of the old families than 
the first members of families that were started later. Suppose 
that of the families starting in 1825, one in ten contained a 
twelfth child, which lived for 60 years. Suppose also that of the 
families starting in 1875 only one in one hundred had a twelfth 
