227 
we should be led utterly astray, as nearly all those names 
are foreign. 
4. As regards the requisite number of cases, a few that are 
wisely selected and accurately reported are better than very 
many that are not. Each error sacrifices several good 
observations before it is diluted until it disappears ; but if a 
faulty bias (as in par. 3) runs through all the observations 
no increase in their number will eliminate it. Otherwise, the 
rule is that the precision varies as the square root of the 
number of observations ; thus, twice the precision necessitates 
four times the labour. It is the best plan to proceed tenta¬ 
tively ; if the results fall into more harmonious sequence as 
you proceed, it is worth proceeding ; and if after dividing 
your statistics into 2, 3, or 4 groups you find the groups agree 
pretty well, and that their sums form a yet more regular 
curve than that obtained from any of the subdivisions, you 
may safely trust it. 
5. Variability.—Mention is above made of “ homogene¬ 
ous” groups : this epithet is applicable when individual 
differences are entirely due to the aggregate effect of a 
great many small and independent variable influences. Ex. 
The stature of an English male adult is due to his being a man 
of English race, reared under the range of those conditions 
of food, temperature, clothing, disease, and the like which 
prevail in England. The large causes common to all are the 
English breed and the range of English conditions ; the small 
causes are differences of varieties and families, and of food, 
temperature, clothing, and the rest, within the range. Varia¬ 
bility depends wholly on the fact of multifariousness of 
causation , and is subject to the well-known law of devia¬ 
tion, which has no more to do with the particular items of 
multifariousness than the rules of arithmetic have to do with 
the quality of the things to be added or multiplied. Two and 
three make five, whether the objects be pence, or peas, or bills 
before Parliament; so the law of deviation holds for the 
stature of men and animals, and apparently, in a useful degree, 
for every homogeneous group of qualities or compound quali¬ 
ties, mental or bodily, that can be named. It is a very general 
Q 2 
