ON THE INFLUENCE OF WEATHER UPON MORTALITY. 305 



There remains, then, only one more subject for preliminary consideration, and 

 that is, the sources of the different data employed in the following pages. The 

 meteorological data are taken from the collected returns from the stations of the 

 Meteorological Society of Scotland, as reduced by the Astronomer-royal. The 

 stations of the Society have a mean latitude of 50° 30' N. ; a mean longitude of 

 3° 4f W. ; and a mean elevation of :222 feet nearly. It will be observed that the 

 meteorological data are deficient in two points, namely, concerning electricity and 

 ozone. Unfortunately, I have no means of applying these subtile agencies to the 

 mortality of the years under examination ; with respect to electricity, indeed, I 

 have no information whatever ; and concerning ozone, I have nothing trust- worthy. 

 It is true that the Meteorological Society's reports contain the results of observations 

 made with the usual test-papers in different parts of the country ; but I submit, 

 with all deference, that until the chemistry of ozone is more fully understood, its 

 physiological action cannot be accurately defined. So long as it is left to each ob- 

 server to determine the amount of ozone present at his station by the varying depth 

 of colour on a slip of paper, our knowledge of the true quantity present must depend 

 upon very slender evidence, and consequently be of very questionable accuracy. 

 It is quite possible that six different observers might, with exactly the same indi- 

 cation on the test-paper, refer the amount of ozone present to as many different 

 shades on the reference paper. Whether the paper itself affords a true indication 

 of the presence of ozone, and to what extent, in the atmosphere, is a disputable 

 matter. At all events, under existing circumstances, I should hesitate in com- 

 paring the ozone returns with the death-rate. 



With respect to the humidity or dryness of the atmosphere, I have employed 

 only three columns, showing, respectively, the number of rainy days, the amount of 

 rain in inches, and the degree of saturation, as deduced by Mr Glaisher : full satu- 

 ration = 100. I have therefore omitted the readings of the dry and wet bulb thermo- 

 meters, the temperature of the dew-point, and the elastic force of vapour. I have also 

 omitted from the tables, although plotted in the diagrams, the absolute highest and 

 absolute lowest temperatures at any of the stations ; these are exceptional records, 

 and can have no general application to the subject of the present inquiry. 



The mortality tables are constructed from the returns made by the Registrar- 

 General. The period over which my investigations extend is six years, namely, 

 from 1857 to 1862 inclusive. This, I conceive, is quite long enough to indicate the 

 relationship existing between the weather and mortality in non-epidemic years. 

 I would, however, have made the period seven years, by including 1856, but I found 

 that the meteorological data for that year were not trust- worthy ; a circumstance 

 arising from the newness of the Society, the inexperience of the observers at many 

 ' of the stations, and a want of proper correction for the errors of the instruments 

 then employed. 



The absolute facts concerning the meteorology and mortality of the seventy- 



