1 p.m. 



Range. 



155 in. 



•033 in. 



301 ,, 



•034 „ 



150 Transactions of the South African PJiilosoj^hical Society. 



upon the mean dryness of the atmosphere, but upon the diurnal 

 variation of vapour tension as computed by the psychrometer. He 

 compared the diurnal variations of vapour tension at Nertchinsk and 

 Makerstoun in 1844, which were as follows : — 



4-5 a.m. 



Nertchinsk -122 in. 



Makerstoun -267 ,, 



As the range is as great at one place as at the other, there can be no 

 better reason (so far as this point is concerned) for the barometric 

 oscillation being single at Nertchinsk than at Makerstoun." 



The whole of Broun 's remarks deserve more attention than they 

 seem to have received. He had, indeed, in 1846, maintained the 

 insufficiency of Dove's method, being in this supported (circa 1852), 

 to some extent by Guyot.''' 



The constants in the Bessel Series for the monthly values of the 

 diurnal curves of pressure are given in Table 28. In the main they 

 are on much the same lines as physicists have determined elsewhere. 

 Perhaps, since they depend only upon four years of observation, it 

 would be unfair to draw very hard-and-fast conclusions from them. 



The angle Vi has a maximum in September and October, and 

 a minimum in June. The amplitude ?ti has, upon the whole, a 

 maximum at midsummer and a minimum at midwinter. 



The angle V2 has minima in February and July, and maxima in 

 October and June. In this it agrees very closely with those made 

 out from both theory and observation by Hann, for Jaluit, in the 

 Marshall Islands, within six degrees of the equator. The monthly 

 values of the amplitude il^ have the cosmical characteristics indicated 

 by Hann : namely that the maxima occur at the equinoxes, and the 

 minima at the solstices ; and that its magnitude is greater at peri- 

 helion than it is at aphelion. The greater maximum at Kimberley, 

 however, is in September, whereas in India it is in March ; f so that 

 it would seem to be not quite independent of the earth's seasons. 



The angle V3 is for half the year in the vicinity of the zero-point, 

 and in the vicinity of 180° during the other half. " Cole points out 

 that at all stations the phase-time is reversed at the equinoxes." j 

 At Kimberley the reversal is pretty rapid at the vernal equinox, but 

 at the autumnal equinox it is much more gradual. The amplitude 



* See the Smithsonian Meteorological TahlcK, p. xxiii, 1893. Also Scott,, 

 Kleinentanj ISIeteorolocjy , p. 110. 



f See Eliot, Indian Met. Monulrs, vol. xii., part 2, p. 286. 

 J Quavtcrlij Journal of the R. Met. S., vol. xxv. , p. 52. 



