X44 



NATURE 



'^Juae 1 6, 1 88 1 



showing 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 cal- 



culated in percentages, which, when plus, arc placed above 

 the mean line of the figure, and when minus, below it. 

 Thus as regards scarlatina (Fig. \), the mean of 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 fro.n small-po.\. 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 corresponding to 



the minimum period of the other, and vice versA. It is 

 also seen that the mortality curve for small-pox (Fig. 3) 

 is quite distinct from the other two curves. 



In order to ascertain the degree of steadiness of these 

 curves, a curve was calculated and drawn for each of 

 the seven epidemics of scarlatina and for each of the 



