Diurnal Meteorological Conditions at Kimherley. 137 



ever, nearly 4° greater, falling 3° lower at night and rising 1° higher 

 by day. It is a curious fact, in this connection, that when the sky is 

 clear the hourly temperature values exhibit an upward tendency, so 

 that in nearly every month the temperature is higher at the close of 

 the day than it was at the beginning. 



Turning now to the hourly variation of temperature on clear days 

 as given in Table 6, we find that while the instants at which tem- 

 perature is changing most rapidly, rising or falling, are almost 

 identical with those of Table 4, yet XV. is the invariable hour at 

 which the hourly temperatures are highest. Considering the hourly 

 values alone, that is, the time of maximum shows a tendency to 

 symmetry with the times of sunrise and sunset, both. We conclude 

 that the maximum temperature does not depend so much upon the 

 length of the day as upon the altitude of the sun, the undisturbed 

 temperature rising not for a fixed time after the sun's appearance 

 above the horizon, but for a fixed time after that has reached its 

 greatest meridian altitude. 



A closer examination of Table 5 will tell us that the actual maxi- 

 mum temperature does not always fall exactly at XV., but either a 

 little before or a little after. The fact is pretty obvious from a 

 consideration of the sequences of hourly values, without going 

 into any refinements of analysis to determine what the actual 

 epochs are. It is a curious fact, however, that if we were to apply 

 a correction for the equation of time, so as to reduce the observed 

 temperatures to their equivalents at hours of apparent time — in other 

 words, referring them directly to the sun — the resulting maxima would 

 approximate even more closely to a given hour of the day. Taking the 

 equation of time into account, and remembering that Cape Colony 

 mean time for the meridian of 22° 30' East is used, while the 

 longitude is really 24° 40', it appears that the true apparent time 

 of maximum is very near to XV. And since this is the hour at all 

 seasons, irrespective of the sun's distance and meridian altitude, it 

 seems to follow that it is also a planetary time, or, in other words, 

 the hour of maximum temperature on any planet in w^hose atmo- 

 sphere clouds do not form. 



The deviations from the normal of temperature on clear days, hour 

 by hour, are git en in Table 7; the hourly values in each month 

 having first been increased or decreased by a numerical constant 

 which has brought the monthly means of the clear days into coinci- 

 dence with the normal means. Heavier type is used to indicate that 

 the deviation is ijIus ; in all other cases it is minus. It is evident 

 that the temperature on clear days is longer above the normal in 

 summer, and the magnitude of the deviation greater than it is in 



