Surron—Influence of Water-Vapour upon Nocturnal Radiation. 25 
by the introduction of error from this source, monthly averages 
have been formed for clear nights only, not a single night being 
included in which any cloud at all could be seen. They are then 
arranged in two sets, one being a sequence of dew-points, the other 
a sequence of relative humidity :— 
MontHty AVERAGES ARRANGED IN OrpDER oF DEw-Pornt. 
Average Dew-Point | Relative Wind- Fall of 
Monru. Number of at Humidity at | Velocity, |Temperature, 
Clear Nights. 8 p.m. § p.m. 8-9 p.m. 8-9 p.m. 
A Per cent. M. per H. B 
Webs s 7 48°8 43 4-8 2°0 
Mar:,. 10 47-5 50 4°7 2-2 
April, . 12 46-4 61 4+] 1°7 
Dec. 2 9 44°1 Oo” 61 2°4 
ame Osi. 10 42°6 34 6:3 2°5 
INOWAne sos 15 37°9 34 6-2 2-6 
Wile ny 85 IE 20 37°0 56 3°8 1:8 
Oct 12 36°9 40 5°6 2°3 
Sept. =. 15 36°8 46 4-4 2°3 
Jane, . 20 32°8 61 4-4 1-6 
DLV ap ae, . = 22 32°4 58 3°9 1-9 
ieee 20 32°2 47 4-3 2-1 
1 It is important that the sky should be absolutely clear, because the least trace: 
of cloud, even near the horizon, suggests condensation in the vicinity. 
[TasLe oF Retative Humipiry. 
