342 Transactions of the Boyal Society of South Africa. 
For each given month the month's mean values of each of these 
elements are arranged according to the magnitude of the range of tem- 
perature in any year 1897 to 1910, some months of 1911 having been 
included. The method of arrangement will be followed in Table 1, 
showing the results for one month, i.e., July, of which it is perhaps only 
necessary to say that the minus sign in the dew-point column indicates 
that the vapour pressure is less at VIII. o'clock than it is at XIV. ; the 
plus sign when it occurs — in other months — that the vapour pressure is 
greater at XIV. 
Table 2 gives for the year corresponding information to that which 
Table 1 gives for the month ; the " A " line, of the year, containing for 
each column the averages of the A " lines of each month; and so on 
for the " B," " C," "D," ... lines. The numbers in each column of 
Table 2 have, however, first of all been smoothed in threes by Bloxam's 
method. 
Table 2 would seem to be of peculiar interest. With one exception all 
the elements in question respond to the rise and fall of the range of 
temperature curve. The following increase more or less regularly as the 
mean monthly range of temperature increases : — 
2. The maximum temperature ; 
4. The mean temperature ; 
9. The range of pressure ; 
11. The duration of sunshine. 
The follow^ing decrease more or less regularly as the range of tem- 
perature increases : — 
3. The minimum temperature ; 
5. The range of the dew-point ; 
6. The mean dew-point ; 
7. The range of relative humidity ; 
8. The mean humidity ; 
12. The cloudiness of the sky ; 
13. The mean rainfall ; 
14. The number of rainy days in a month. 
The one exception is the mean pressure of the month, which com- 
mences by falling, but concludes by rising as the range of temperature 
increases in magnitude. Whether this anomalous behaviour of the 
