76 
(Public Health, p. 87) appears to have observed approxi- 
mately a ten or fifteen years’ period. But a little closer 
examination will probably account for the neglect to notice 
the recurrence of these events. The life of any one physician 
would be too short to admit of the observation of more than 
four or five such visitations, and he might well be excused 
if he hesitated to draw any conclusion from so small a basis. 
Intercommunication between medical men was also more 
difficult than it has become now-a-days. 
No numerical records of the causes of death were then 
kept for subsequent scrutiny, and it is probable that many 
of the less fatal forms of these diseases would pass through 
a nation without much notice. 
Even when there is a registration of sickness as well as 
of deaths it does not always prove possible to mark the 
occurrence of an epidemic period. Quite recently, in 1868, 
Dr. Ballard, in an analysis of records of sickness kept in 
Islington for 12 years, was unable to find any indication of 
the cyclical recurrence of whooping cough, though he 
assigns a period of two years to measles, and four to scarlet 
fever. 
It was, however, this disease, whooping cough, that first 
impressed upon my mind the fact that epidemic cycles do 
exist. 
From the returns of sickness published every week by 
the Manchester and Salford Sanitary Association from the 
year 1860, it was evident that whooping cough tended to 
reappear biennially. Every second year the curve of this 
disease rose in the colder months of the year — in most 
cases epidemically. The rise and fall of the wave is almost 
perfectly regular. I have constructed a diagram that 
displays its course graphically, and it will be seen from 
this that whilst there are comparatively few cases in the 
winters of 1860, ’62, ’64, ’66, ’68, ’70, ’72, ’74, ’76, and ’78, 
in the cold seasons of the intervening years tliere was 
