[ 796 ] 
will come out 720180 and 7765 20 ; of which the 
difference is 56340, the number decreafed for the 
lad five years. 
There is another way of computing from the 
number of houfes ; but I think this not fo certain 2s 
the other. For here are two difficulties, to afeer- 
tain the number of houfes, and to fix on the num- 
ber of perlons for each houfe. As to the lad, Sir 
William Petty thought we might allow eight perfons 
to a houfe ; which I have found to be a midake. I 
have made an experiment of it, and got an exadl 
account of the numbers in each houfe in a certain 
paridi in London j and I find that they exactly come 
to fix in a houfe, empty and full together, for there 
is feldom above one in twenty empty. And as in 
that paridi the people are in a middle condition, and 
fome of them have a number of fervants ; it may be 
prefumed they are in a middle date with regard to 
numbers, between the very great families and thofe 
in the lowed rank. This is alfo confirmed, if we 
allow, as above, one in thirty to die yearly in Lon- 
don. For within the city walls there were 11857 
houfes in the p7 parifhes, as appears from Mr. 
Smart’s account, which was fuppofed to be very 
accurate at that time : But flnce he publidied 
that in 1741, there are not fo many houfes within 
the city walls ; for in many parifhes there are houfes 
greatly enlarged, fome rebuilt in place of two or 
three, and warehoufes made of others. I know 
fome parifhes in which there is one in twenty fewer 
than in his time. In others perhaps there is no al- 
teration . But I think they mud, at an average, be 
diminifhed three in an hundred at lead j and confe- 
quently 
