MISCELLANEA. 
On the Distribution of Severity of Attack in Cases of Smallpox. 
By F. M. turner, M.I). 
On Vol. IV. pp. 505-510 of Biometrika, Prof. Pearson give.s reasons for believing that the 
distribution of severity among cases of smallpox is either normal, or not sufficiently skew to 
sensibly affect the calculations of correlation tables by normal curve formulae. His arguments 
are partly a priori ; of direct evidence he only produces a list of cases of smallpox classified 
according to the length of time the patients were considered too ill to be bathed. 
The following evidence to the contrary seems to me very strong. For over 10 years 
Dr Ricketts, formerly the Superintendent of the Hospital Ships, now Superintendent of all the 
smallpox hospitals of the Metropolitan Asylums Board, has divided his cases into six classes 
defined as 'follows : 
Class I. Haemorrhagic cases. 
„ II. Cases confluent in the vesicular stage. 
„ III. Cases confluent, but not before the pustular stage. 
„ IV. Cases intermediate between classes III and V. 
„ V. Cases with from 100 to 500 pocks upon the face. 
„ VI. Cases with less than 100 pocks on the f\ice. 
It will be seen that all the classes except I are defined by the severity of the eruption. In 
Class I the cases are so severe as to die, almost without exception, either before any eruption 
appears or before it is fully developed. It is almost certain that the cases of this class would 
belong to either Class II or III, if they lived long enough. 
When I was working under Dr Ricketts two years ago, he generously put his records at my 
disposal, and I found that in the year 1902 the following number of patients were treated at 
Long Reach Hospital and the Hospital Ships : 
Class I. 266 Class IV. 1141 
II. 291 „ V. 1385 
„ III. 1015 „ VI. 2851 
Total 6949. 
To get the comparative frequency of different degrees of severity of eruption we require a 
quantitative definition of the classes, which is given in Classes V and VI and may be ascertained 
with some approach to accuracy, as I shall shew below, in Classes II, III, and IV. That severity 
of disease in smallpox is very closely connected with severity of eruption is quite evident to me ; 
and will, I believe, be admitted to be so by all who have had practical acquaintance of the 
disease. 
To determine the superior limit of Class IV Dr Ricketts has sent nie a photograph taken by 
his assistant, Dr Byles, which has been used as a standard, dividing Classes 111 and IV. The 
photograph is endorsed " Class III. Cases of less numerical severity fall into Class IV." The 
photograph is a profile view and on the half face I have counted 672 pocks. This number must 
not be taken as absolutely correct, partly because it is difficult to distinguish the individual 
pocks which have run together to form a group, partly because pocks are found of all sizes ; 
besides those fully formed are others of small size, and others abortive, and it is difficult to draw 
a definite line in counting. Still the evidence shows that the division between Classes III and 
IV corresponds to about 1300 pocks. 
