84 Transactions of the South African Philosophical Society. 



when the sky is half-clouded. This seems to indicate, what is other- 

 wise not improbable, that when the sky is more than half covered 

 the clouds are as likely to shut off the solar heat as to impede 

 radiation from the thermometer. Under certain circumstances, more- 

 over, clouds may reflect heat to the thermometer. The temperature 

 in the shade seems not to be so much influenced by the amount of 

 cloud. It falls off a little when the percentage of cloud exceeds 60 per 

 cent., and notably so when the sky is quite overcast. Perhaps on 

 the whole the greatest cloud effect upon the temperature of the lower 

 air is somewhere about 40 per cent. The difference of maxima 

 between sun and shade, however, goes on increasing up to a cloudi- 

 ness of 70 per cent, or 80 per cent. The explanation of this fact 

 seems to be that a clouded sky accelerates the time of maximum 

 shade temperature, changing it from the normal at 3 p.m. when the 

 sky is clear, to 1 p.m., or earlier, when there is much cloud. Thus 

 the shade temperature will not go on increasing for so long a time 

 under a very clouded sky ; consequently the rise, after the time 

 when the black bulb has attained its maximum, will be less in 

 magnitude. There are rare occasions, in cloudy weather, when the 

 maxima in the sun occur late in the afternoon. 



TABLE IV. 



The Elements aeeanged in a Sequence of Cloud 



Peecentages. 





Maxima in the Sun. 



Difference of 

 Maxima. 



Dew-point at 

 Noon. 



Eelative Humidity 

 at Noon. 



Cloud 



























%• 



Jan. 



May 



Sept. 



Jan. 



May 



Sept. 



Jan. 



May 



Sept. 



Jan. 



May 



Sept. 





to 



to 



to 



to 



to 



to 



to 



to 



to 



to 



to 



to 





April. 



Aug. 



Dec. 



April. 



Aug. 



Dec. 



April. 



Aug. 



Dec. 



April. 



Aug. 



Dec. 

















o 























% 



% 



% 



. . . 



144 



124 



141 



58 



54 



59 



43 



37 



40 



28 



36 



26 







145 



125 



144 



58 



54 



59 



47 



38 



43 



30 



37 



27 



10 



145 



124 



142 



59 



54 



59 



50 



39 



43 



35 



38 



28 



20 



147 



123 



145 



60 



55 



60 



50 



38 



43 



35 



39 



28 



30 



148 



125 



146 



61 



54 



60 



51 



39 



45 



36 



39 



29 



40 



150 



125 



147 



63 



56 



63 



52 



41 



44 



37 



41 



30 



50 



150 



125 



149 



64 



56 



64 



53 



37 



45 



37 



38 



31 



60 



150 



125 



147 



65 



57 



64 



54 



40 



45 



41 



44 



32 



70 



148 



122 



147 



65 



57 



64 



54 



42 



45 



42 



49 



31 



80 



149 



120 



148 



66 



56 



65 



55 



37 



47 



45 



45 



33 



90 



137 



122 



136 



59 



53 



57 



56 



42 



50 



60 



46 



50 



100 



126 



88 



125 



52 



31 



50 



58 



41 



53 



67 



62 



55 



Table IV. gives a subdivision of the elements of Table III. into 

 periods of four months each. The most striking fact is that during 



