Ch. ix. in England (continued). 427 



the diminished numbers of 1796 and 1800, was 

 more likely to be below than above the general 

 average. For these reasons, together with the 

 general impression on the subject, it is probable 

 that the enumeration in 1800 was short of the 

 truth, and perhaps the population at that time 

 may be safely taken at as much as 9,287,000 at 

 the least, or about 119,000 greater than the re- 

 turns gave it. 



But even upon this supposition, neither the 

 excess of births above the deaths in the whole of 

 the ten years, nor the proportion of births to 

 deaths, as given in the registers, will account for 

 an increase from 9,287,000, to 10,488,000. Yet it 

 is not probable that the increase has been much 

 less than is shewn by the proportion of the births 

 at the two periods. Some allowance must there- 

 fore necessarily be made for omissions in the 

 registers of births and deaths, which are known 

 to be very far from correct, particularly the 

 registers of births. 



There is reason to believe that there are few 

 or no omissions in the register of marriages ; and 

 if we suppose the omissions in the births to be 

 one-6th, this will preserve a proportion of the 

 births to the marriages as 4 to 1, a proportion 

 which appears to be satisfactorily established 

 upon other grounds ;* but if we are warranted in 

 this supposition, it will be fair to take the omis- 



* Sec the Preliminary Observations on the Population Abstracts, 

 p. xxvi. 



