1867.] 



Insolation to Atmospheric Humidity. 



363 



pour, those nights being considered clear in which the percentage of cloud 

 did not exceed -10. It was found, on a careful tabulation of hourly obser- 

 vations, that the fall in temperature was decidedly greater when the quantity 

 of vapour was relatively small*. 



Cloud, haze, and opalescence t of the atmosphere more probably the 

 principal cause of the phenomenon. — It appeared of much importance to 

 ascertain whether the presence of cloud, and the imperfect state of trans- 

 parency in the sky which usually accompanies it, may not have materi- 

 ally assisted in producing the results alluded to in the last paragraph, and, 

 u fortiori, account for the increased insolation noticed in cloudy weather in 

 various parts of India — e. g. " on days in the rainy season when the clouds 

 are temporarily broken," and, as in Sikkim, "when a break in the clouds 

 of an hour or two had to be watched for to obtain observations of solar ra- 

 diation " (Schlagintweit, ' Meteorology of India,' pp. 49 & 51). 



To test this point, the results of the observations for the four years end- 

 ing 1844, over which the inquiry at Madras extended, were divided into 

 groups of contemporaneous observations of temperature, vapour-tension, and 

 percentage of clear sky, when the mean numerical results showed that a 

 progressive fall of about 1° F. for every "10 of vapour-tension was accom- 

 panied by a proportionate increase in the percentage of clear sky, — a result 

 which is the more significant when it is considered that the infusion of 

 visible cloud was limited to '10. 



To prevent mistake, the fall in temperature on the nights of maximum 

 clearness was compared with the fall on nights of minimum clearness 

 within the limits above stated. There were twenty-five nights which were 

 estimated to be perfectly clear, and twenty-two nights when the per. 

 centage of clear sky ranged from "90 to '93, the average being '915 (1*000 

 representing an entirely clear sky). The results were as follows : — 



The mean fall of temperature at an estimated clearness of sky denoted 

 by 1-000 was 8°-3 F. 



The mean fall of temperature at an estimated clearness denoted by '915 

 was F. 



The diiference is 2°'3. The contemporaneous mean tensions of vapour 

 were '08 and '83 respectively. 



It is probable then, as cloud accompanies humidity, that tension of vapour 

 gives some indication oiihQ state of transparency of the sky, yihW&i afford- 

 ing a measure of the quantit}' of invisible vapour in the air. 



In any case it is sufficiently clear, both from the results at Madras and 

 from the slight increase in the tension at 2*^ p.m., that the amount of 

 aqueous vapour alone is not sufficient to account for increased or diminished 

 solar radiation. 



" Pliil. Mag. October 18C6 ; where see Tables and method of deduction. 



t This term was first used by Professor Eoscoe. It here represents tlie state of the 

 atmosphere at the moment vapour is in process of condensation previously to its forma- 

 tion into cloud. 



