6 S. A. Hill — The Measurement of Solar Uadiation. [No. 1, 



nine equations, from which, by the method of least squares, I have arrived 

 at the following results : 



E = 80-° 434* 



a = -99856 



;8 = -71186 



y = -99393 

 The value of a is identical with that which I obtained from Mr. 

 Hennessey's observations at Mussoorie, while that of y8 is intermediate 

 between the two values computed from the observations of the 12th and 

 14th November 1879, at Mussoorie. f Clear dry air of 30 inches pressure 

 absorbs about 4i per cent, of the total radiation which falls upon it verti- 

 cally ; while water vapour, with a pressure of 1 inch below and the average 

 vertical distribution, absorbs 29*8 per cent. The efEect of dust is less than 

 might be expected, the loss due to this cause in May being apparently 

 only about 6 per cent. ; but this is doubtless because the dust is not a 

 simple absorbent, for it scatters or reflects the rays in all directions, and 

 some of these reflected rays reach the globular bulb of the thermometer. 

 An actinometer, arranged to receive parallel rays only, would indicate a much 

 greater loss, on account of suspended matter in the atmosphere. 



The monthly means computed by these constants, and their variations 

 from the observed means are as follows ; — 



Jan. Feb. Mar. Apl. May June Oct. Nov. Dec. 

 62-9° 64-0° 63-4° 629° 59 6° 57'4<' 58-6° 64-4» 62 2° 

 —0-7 +0-2 +0-8 +1-3 —0-9 —2-2 +0*2 +1-1 +0-2 



The most important difference is in June, when the computed value is 

 2*2° in defect. This is almost certainly the result of taking the mean 

 vapour tension for the whole month, instead of that for the clear days only 

 The other differences are probably due, in great part, to the unavoidable 

 neglect of variations in the condition of the ground surface. 



Applying the formula to the observed mean radiation temperatures 

 for the nine dry months of each year, and taking the proportionate number 

 for dust to be the same each year, we arrive at the following results : — 



Year 



Corrected radiation 



temperature. 



The last minimum of sun spots having occurred in 1878, while 1882 

 was probably a maximum year, these figures indicate a variation in the 

 solar heat of very considerable range, and in the opposite direction to that 



* The degrees of the sun thermometer are only arbitrary units, which cannot be 

 readily converted into units of heat per unit of surface per unit of time. 

 t Proceedings of the Royal Society, No. 219, page 436. 



1876 



1877 



1878 



1879 



1880 



1881 



1882 



80-05° 



82-60° 



82-67'» 



81-23° 



8023° 



79-12° 



7711* 



