588 SEDCJWICK AND WINSLOW. — BACILLUS OF TYPHOID FEVER 
■ 
The mean moiitlily temperatures for the American cities were obtained from the 
rcpoi-ts of the United States Weather Barean; those for the German cities, from 
ublicatlons of the astronomical observatories in their resp 
d 
those for London, Paris, Montreal, Buenos Ajres, and Santiago from special local 
piiMIcaiions mentioned in connection with the tables. For the States of New York 
and Massachusetts, it was assumed that the temperature of New York Citj and Boston 
would serve without serious error. For Japan, where the range of temperature is 
rather wide, an average was taken of the record of ten stations in different parts of 
the Empire, as given by the Central Meteorological Observatory. In the case of India, 
it appeared inadvisable to attempt to calculate an average for the whole empire, as the 
seasons in the difTerent districts are so very different. The typhoid figures are, there- 
fore, compared with two sets of temperature values, for Central India, and for the 
Punjab, taken from lUnrC^ '' EUmatohgie,'' which give a fair idea of the two most 
imi^)()rtant meteorological zones. For each of the cities and stations, with one or two 
exceptions, the figures for ten years have been used in order to secure a reliable 
average; and the mean monthly temperatures finally obtained have all been reduced 
to the Fahrenheit scale for uniformity and convenience in plotting the curves. 
The typhoid statistics include records of hospital admissions at the two hospitals 
of Santiago de Chile, of hospital admissions in the British Army in India, of reported 
cases at Newark and of deaths in all other instances. The figures for the American 
Stiites and cities, for Montreal, London, and Paris, were obtained from the published 
reports of the local Departments of Health, supplemented in some cases by informa- 
tion furnished in reply to correspondence; the German statistics were taken from the 
''Verofcnilichmrjm des Kmserlichen GcsundhcUsamtes ;'' for Japan, the Annual Reports 
of the Central Sanitary Bureau, for India, the Parliamentary blue-books, and for the 
South American cities, local sanitary periodicals referred to in the tables, were cbn- 
sulted. The figures for ten years were averaged in each case except as follows: for 
Vienna and Japan the period was five years; for Atlanta, six years; for Montreal 
and New Orleans, eight years; for Denver and Paris, nine years; for the Army in 
India, eleven years; for Buenos Ayres, twenty-two years. In each case the average 
number of deaths per month has been reduced to a ratio of one hundred deaths per 
year, the final figure for each month representing the number that occur in that 
month for every hundred deaths in the year. Thus the absolute amoimt of the 
disease is entirely eliminated, and only its seasonal distribution considered. The value 
of the statistics will not therefore be impaired by errors of registration, which it may 
be assumed will not vary from month to month. 
i 
