222 



Mr. S. A. Hill. On the Constituent of [Dec. 22, 



dry air and water vapour respectively, taking -L— ^ D ^°^' ^ M ; 

 ^_l Q r^/i)~l°g /m w here/ i s the vapour tension, and B the height 



C 



Ah 



of the barometer corrected for vapour tension, we get the following 

 values : — 



Date. 



C 





C 2 



4th November, 1869 



12th „ 1879 



14th „ „ 



19,922 

 26,213 

 24,840 



64,409 

 66,370 

 66,066 



17,210* 



26,814 



25,499 





23,658 



65,615 



23,174 



In computing C 2 the vapour tensions observed at the Meteorological 

 Observatory of Roorkee have been combined with those of Dehra, 

 because the latter seem to be somewhat too great at all times of the 

 year in comparison with those for other places in the Himalaya at 

 nearly the same altitude, f while Roorkee is generally a little too dry. 

 In any case, with such a variable element as water vapour, it is best 

 to deduce the law of distribution from as many observations as 

 possible. 



The difference between the values of C and C 2 is so small, con- 

 sidering the nature of the observations, while that between C and Ci 

 is so very great, that there can be very little error in agreeing with 

 Dr. Tyndall that the absorptive power of dry air is sensibly nothing, 

 and that the total absorptive power of the atmosphere is due to the 

 water vapour it contains. 



The true absorption coefficient at sea-level, on the supposition that 

 the absorbent substance increases in density according to the same 

 law as between Mussooree and Dehra, appears to have been a good 

 deal less on the two days of observation in 1879 than the mean value 

 found by Pouillet in 1838, while its value on the 4th November, 

 1869, differed but little from Pouillet's mean result. Pouillet found 

 the absorption of a vertical column of transparent atmosphere at 

 Paris to vary between *18 and '27 of the total incident radiation, the 

 mean being about *24. From Mr. Hennessey's observations the co- 

 efficient for Naperian logarithms at sea-level appears to have been 



* Computed from the observations made at the Meteorological Observatories of 

 Dehra and Eoorkee, and at Chakrata, 30 miles north of Mussooree and 7,160 feet 

 above the sea. No observations of humidity were made at Mussooree on this day. 



f This has been pointed out by the writer in a paper on the meteorology of the 

 North West Himalaya, published in the " Indian Meteorological Memoirs," vol. i. 



