630 



Mr. Alexander Buchan 



[March 25, 



mortality from which is largely determined by season and weather, a 

 comparatively small number of years is required to give a satisfactory 

 approximation to their true weekly curve of mortality. But as 

 regards the great majority of diseases, it quickly became apparent that 

 a thirty years' average was required in the construction of curves 

 which could be offered as true " constants " for the diseases to which 

 they refer. The thirty years beginning with 1845 were therefore 

 adopted. An examination of the curves shows that some of their 

 striking features, particularly those indicating the complications of 

 special diseases and their connections with each other, which the 

 weekly averages disclose, would entirely disappear if monthly averages 

 only were employed. 



The curves of the more prominent and interesting of the diseases 

 are shown on the accompanying woodcuts, the straight black line in 

 each figure being drawn to represent the mean weekly death-rate on 

 an average of the fifty-two weeks of the year, and the figures on the 

 margin the percentages above or below the average. With this 

 general average the mean death-rate of each week is compared and 

 the difference above or below calculated in percentages, which, when 

 plus, are placed above the mean line of the figure, and when minus, 

 below it. Thus as regards scarlatina (Fig. 1), the mean mortality of 



Fig. 1. 



Moan ^ 



Scarlet Fever. 



the fifty-two weeks is 49 * 6 ; on the first week of January it is 7 per cent, 

 above the mean, from which time it continues to fall to the annual 

 minimum, 35 per cent, below the mean in the middle of March ; thence 

 rises to the mean in the end of August ; to the annual maximum, 60 

 per cent, above the mean, in the end of October, and thereafter 

 steadily falls. The portion of the curve above the mean line thus 

 shows the time of the year when, and the degree to which, the 

 mortality from scarlatina is above its average, and the portion below 

 the line when it is under it. 



Fig, 2 shows similarly the distribution of the mortality from 

 whooping-cough through the weeks of the year, and Fig. 3 the 

 distribution of the mortality from small-pox. It is seen at once that 

 the mortality curve from scarlatina is precisely the reverse of the 

 curve of whooping-cough, the maximum death-rate period of the one 



