110 Proceedings of Royal Society of Edinburgh. [sess. 
result of this investigation is that, when the cumulative variability 
of temperature did not exceed 14°, the weekly number of deaths 
amounted to only 16 ‘5, but when the variability was above 28° the 
deaths numbered 22*5, an increase of 36 per cent. The following 
are the values : — 
Below. Above. 
Variability, . . 15° 15°-21° 22°-28° 28° 
Deaths per week, . 16’5 18*7 19*9 22‘5 
It has been long known that the mortality from the three 
diseases under review varies inversely as the temperature. This 
applies to all but the lowest temperatures, as will be apparent from 
the following table, the returns being for the ten years specified 
above. 
Below. Above. 
Temperature, . 36° 36°-40° 41°-45° 46°-50° 51°-55° 56°-60° 60° 
Deaths per week, 25 '3 26‘0 22’4 17*4 13‘5 13T 11'6 
The lowest mortality from bronchitis, pneumonia, and pleurisy 
occurs at the end of August, after a period of high temperature and 
small variability. 
Tig. 11. — Fever. 
The returns include deaths from the following fevers, viz., 
typhoid, typhus, relapsing, cerebro-spinal, simple and ill-defined. 
The mortality * from the latter diseases is so small that the deaths 
from them may be neglected. The usual autumnal rise of typhoid 
is well marked, the maximum fatality occurring at the end of 
November. 
Small-Pox. 
Nearly all the cases of small-pox having occurred during only one 
epidemic, that of 1894, it is not possible to construct a curve for 
this disease, but half of the deaths recorded took place in December 
of that year. 
Violence. 
The curve of deaths from violence is extremely irregular, and 
* The following were the deaths recorded in the county of Edinburgh 
during the year 1893 : — Typhoid, 74 ; typhus, 6 ; relapsing, 0 ; cerebro-spinal, 
0 ; simple, 3. (See Registrar -General’s Detailed Report, pp. 54 and 55.) 
