1895 - 96 .] Mr E. C. Mossman on Death-rate in Edinburgh. 107 
but the mean of the 1st, 2nd, and 3rd weeks, the average of 
the 3rd week being the mean of the 2nd, 3rd, and 4th weeks, 
and so on (see Table II.). The values thus obtained were then 
plotted. It may be stated that a 53rd week had in one or two 
years to be introduced by the Kegistrar-General in order to bring 
his returns into approximate agreement with the calendar. When 
the introduction of a 53rd week was necessary the data given in 
the first weekly return for the year in question were not tabulated. 
Fig. 1. — Deaths of Both Sexes at all Ages and from all Causes. 
The characteristic features of the mortality curve, which is toler- 
ably constant from year to year, are (1) a period of high death-rate 
from about the middle of November to the end of January; and 
(2) a long-continued minimum in summer, extending from the 
beginning of July to the end of September. During the former 
period temperature falls rapidly, with a considerable increase in 
rainfall and humidity. The variability of temperature is at a 
maximum, the general atmospheric conditions being thus raw and 
inclement. During the time characterised by low mortality, air 
temperature is high, humidity low, and temperature variability at a 
minimum. The rapid fall in the death-rate during the four weeks 
ending with the middle of February is of interest, as is the general 
retardation of any further improvement in the mortality returns 
until the end of May, thus confirming the general impression that 
the easterly winds of spring exercise a prejudicial influence on the 
health of the community. 
Fig. 2. — Deaths of Infants under 1 Year of Age. 
Infantile mortality has its primary maximum from November to 
January, a secondary maximum due to diarrhoea being a marked 
feature during the month of August. The minimum extends over 
the six months ending with July, but is not such a pronounced 
characteristic of the curve as the maximum. The absolute mini- 
mum is reached about the end of June, the absence of zymotic and 
bronchial affections at the time being of interest in this con- 
nection. 
