168 REPORT — 1855. 



ral during the fifteen years embraced in the returns, and the consequent alterations 

 and improvements in the returns themselves, the author showed that the year 1845, 

 in which blank forms of certificates of the causes of death were first issued, formed 

 the commencement of an Kra of greatly improved registration ; so that before the 

 year 184S it is reasonable to suppose that the returns of the causes of death would 

 have attained to all the accuracy of which they are susceptible in the hands of the 

 present race of medical men ; and that neither fashion, nor new theories, nor 

 increasing knowledge, would have materially affected them in so short a period of 

 time. In the longer period of fifteen years, even in the absence of statistical instruc- 

 tion, some changes would doubtless have taken place in medical opinions as to the 

 causes of death ; but as these changes would not affect the diseases most easy of 

 diagnosis, such as typhus fever, the eruptive fevers, diarrhoea, &c., the tables would 

 still be found in nianv parts, throughout the whole period embraced in them, to 

 furnish the materials of ajust comparison. 



In carrying out the comparison of year with year, the author proceeded to com- 

 ment upon the several tables forming the Appendix to his paper, in their order, 

 beginning with the births, deaths, and marriages, and ending with the more consi- 

 derable of the special causes of disease; his chief object being to direct attention to the 

 fuciuaiiom in the mortality from special causes, employing as a measure of mean fluc- 

 tuation the quotient obtained by dividing the sum of all the successive differences 

 between year and year, whether in excess or defect, by the number of those differences, 

 and then reducing that quotient to a per-centage proportion of the average of all the 

 years, and also the greatest and least numbers in each series of facts, reducing the 

 difference between them, or in other words, the Extreme Fluctuation, to a per-centage 

 proportion of the maximum numbers. 



Having thus explained the meaning of the terms " Mean Fluctuation," and 

 " Extreme Fluctuation," the paper proceeds to a review of the several tables, making 

 appropricite -comments upon each, beginning with the births, deaths, and marriages. 



The births, which, on an average of the fifteen years, amount to 32,028 in the 

 million, have fluctuated between a minimum of 30,348 and a maximum of 33,736, 

 the mean fluctuation in the intervening period having been nearly 2 per cent. The 

 deaths were subject to much greater fluctuation. They ranged from a minimum of 

 20,925 to a maximum of 30,078, the average being 24,864 ; and the mean fluctuation 

 was 9'51, or little short of 10 per cent. The highest and lowest numbers occurred 

 in the two consecutive years 1849 and 1850. The marriages, fur the shorter period 

 of thirteen years, presented an amount of fluctuation intermediate between that of 

 the births and of the deaths. The average number of marriages was ]0,136 ; the 

 extremes were 9408 and 10,966 ; the mean fluctuation 3'75, and the extreme fluc- 

 tuation 14-20. 



The extreme fluctuations in the numbers of births, deaths, and marriages, follow 

 the same order as the mean fluctuations. Thus the mean and extreme fluctuations 

 in the births were, respectively, r95 and 1004 ; in the marriages 3*75 and 14 20 ; 

 in the deaths 951 and 3038.' 



The births were uniformly in excess of the deaths, and even the fatal year of 1849 

 was no exception to this rule. The excess varied from 1838 in that year to 11,086 

 in the year following. 



From the Table showing the deaths at the three ages to 15, 15 to 60, and 60 and 

 upwards, it appeared that the greatest m^an and extreme fluctuation occurred from 

 15 to 60; the least mean fluctuation from to 15, and the least extreme fluctuation 

 from 60 years of age upwards. The occurrence of the least mean fluctuation in 

 persons under 15 years of age was, perhaps, scarcely to be expected. 



The Table showing the deaths in five districts of the metropolis presented some 

 results worthy of notice. In the east districts alone the maxima and minima coin- 

 cided with the maxima and minima of the total deaths. In the south districts, the 

 maximum number occurred in the same year, 1849, but the minimum number in the 

 year previous, instead of in the year following. In the west and cential districts, 

 again, the minima coincided w-ith the minimum of total deaths ; while the west and 

 north districts were distinguished by the occurrence of the greatest number of deaths 

 in the last year of the series, 1854. 



But the most interesting fact shown by this table was the excessive mortality and 

 high rate of fluctuation prevailing in the southern districts of the metropolis. 



