Ixxxiv Journal of the Asiatic Society of Bengai. [N.S., XIV, 
the same time weddings are numerous especially among the 
working classes, and hence the proportion of those who can 
only put a mark in the marriage register will be large. In 
text-books of statistics there are definite types of relationship 
which are called spurious, but the necessary and sufficient way 
of avoiding all these pitfalls and securing a right result is to 
consider all the factors which can produce an appreciable 
will in general be related almost as clearly as the orig 
series. Hence there is hope that these vital statistics, though 
inaccurate, will yield results which are true qualitatively 
A student of Bengal data finds himself brought up at the 
outset by what seems an insurmountable barrier to per a 
first sight,' appears to be the only interpretation that can be 
put on the close relation between the death-rate and the 
number of chowkidars per 10,000 of the population. But even 
if the work is badly done why should this relation hold? Our 
suspicions are thereby roused, and -if in obedience to the 
that of to-day is high. Tf the number of chowkidars pet 10,0 
is also high in these cases, it means that for the areas 2 high , 
time ; and this, I am told, is what happened. 
curve on the screen gives the ratio of the population 
! See Fig. VIII. 
