- . Dmrnal Meteorological Conditions at Kimberley. 139 



maximum value of the angle {i.e., to a second earliest phase-time) in 

 January or February. Comparing Tables 5 and 8, we shall conclude 

 that these differences owe their origin more to variation in the 

 amount of cloud than to anything else, although the equation of 

 time may be responsible for much of the variation in the monthly 

 values of Vi. 



A comparison between the clear- and the cloudy-day constants is 

 noteworthy. The amplitudes of the former are much larger in each 

 term ; but there is very little difference in the phase-times of V^, Y2, 

 and V4 — twenty minutes being the greatest, in V^. The angle V3, 

 however, is exceptional : not only is the difference of phase-time 

 between the clear and cloudy days greatest of the four, but its affec- 

 tion is also in the other direction, the clear coming earlier than the 

 other. It is indeed curious that the cloud factor should act with the 

 greatest intensity on the term of eight-hours' period, and suggests 

 perchance a cloud-wave oscillating three times a day — shall we say 

 once for each of the three simple cloud-forms ? 



Whether this triple oscillation exists as a general rule is a moot 

 point. Cloud observation is in a very unsatisfactory state. The best 

 published observations of the hourly amount of cloud with which I 

 am acquainted are those of — 



1. A. L. Eotch, in " Observations made at the Blue Hill Meteoro- 

 logical Observatory in the year 1896," pp. 19-27. 



2. The Trevandrum Observations made under the direction of the 

 late J. A. Broun for six days in each week during a number of years, 

 observations on the Sabbath being omitted with a praiseworthy regard 

 for the Shorter Catechism mightily edifying to the Hindoo computers.'" 



3. The hourly observations made at selected stations in India on 

 four days in each month, for ten years or more.f 



In the first, maxima are not improbably indicated at midnight, 

 YIII., and XIV., and the Kenilworth observations, such as they are, 

 agree reasonably with these epochs. The Challenger observations 

 at sea indicate maxima at about VIII. and XVI., from a short series 

 of observations extending only over nine months. At Cordoba, from 

 observations made at irregular intervals during the day, W. G. Davis 

 deduced a formula for the diurnal cloud-period from which he claimed 

 with some confidence two maxima, one about IX. and the other at 

 XVI. ;|: It is curious, though, that of the four specimen months 



* Indian Met. Me>iiolri^, passim. f Ibid. 



:j: "La forma de las curvas resulta bastante irregular, pero demuestra con- 

 cluyentemente la existencia de dos maxima uno por la mafiana y otro en la tarde " 

 {Aiiales de la O^fficma Met. Arg., vol. ix., p. 273). There is, however, an abortive 

 maximum at 1 a.m. 



