466 



STATISTICS. 



[Sect. XV. 



m 



factory computations from governmental departments, and 

 especially those connected with the taxation of the coun- 

 try : but it must be evident that, to render such data 

 available, the eircumsiaiices riudcr which it has been col- 



1 



JL 



ected must; as far as possible, be ascertained and recorded. 

 Where no official accounts cp^ he made available, recourse 



houid be had to private ch?iimels, giving the preference 



f^ 



to siv'h Btatemeiits (if snch exist) as may have been pub- 

 lished in the country, and have thereby been Biibjected to 

 criticism and correction on the part of those best qnahfied 

 to form a judgrnoBt on the subject. Local registers are 

 sometimes to be met witii, where the central government 



has not interfered. Such were carefully kept in many 

 parisliea in En^band, before any government census wab 



midertai 





From 



ncn 



fi rxri 



throusli a series of years, 



cli registers^ comparing births with 



population of a 



country may be estimated with some approach to accuracy. 

 The rate of mortality is a fact of so much importance 

 towards any useful knowledge of a country, that 



it IS 



natural I y am ong th e sub j ects of in qui ry that should 

 earliest command attention. If registers of bui L^loj which 

 record the ages at which the deaths occur, can be obtain^^''. 



they would elucidate many points of ^rreat interest nj> to 



the condition of the people and the effect of the cHmate, 

 a-r 1 would besides afford means, in connexion with die 

 number of births and marriages, for more nearly ap- 

 proximating towards an accurate estimate of the popu- 

 lation. Where a censr- '^^^ been taken, a distinction will 



doT- ^ ties? have been made between the sexes : c...l if the 

 ageft also luwe been recorded, the tahles will themselves 

 afford means for testing their general accuracy, as it may 



b 



ye 



a 



4 



