72 Royal Society. 



thermometer was not covered with an exhausted globe. The mean 

 results of the experiments were as follows : — 



1. A screen J in. in diameter reduced the difference of the read- 

 ings of the black-bulb thermometer and a thermometer in the shade, 

 four yards distant, by one-third. 



2. A screen 2| ins. in diameter reduced the difference by two- 

 thirds. 



On reversing the experiment, converse results were obtained, e. g. 



The rays of the sun, after passing through a circular aperture 

 2| ins. in diameter in a 12-in. screen, were made to fall on the bulb 

 of the solar thermometer, when the readings were found to equal 

 in value those obtained when the instrument was entirely exposed*. 



And no difference was noticed when the black-bulb thermometer 

 was screened from the rest of the sky by a double cover of mill-board 

 placed tent-wise over it. 



Results of an equally negative kind were obtained in the case of 

 other experiments which were made with the object of detecting 

 heat in the light reflected from sky and cloud not in the direction 

 of the sun. 



A black-bulb thermometer, after having been placed for sometime 

 in a dark room, was then exposed to the sky, near a large French 

 window, facing S.E., the glass of which was clear, and had been care- 

 fully cleaned, without any rise being perceptible. The sun, at an 

 altitude of about 40°, was shining brightly on white vapour and light 

 cirro-cumulif. 



Thermometers were also placed in the open air on the north side 

 of the house, on a still day, exposed to half the sky when covered 

 with bright white clouds ; but the mercury stood at the same height 

 as in a dark passage on the same side of the building J. 



The same apparent absence of any direct heating-power in the 

 light reflected from the sky generally was shown in this as in the 

 previous series of experiments when the solar thermometer was 

 screened excepting in the direction of the sun. 



As respects the momentary increase of insolation which occurs in 

 connexion with bright vapour in the neighbourhood of the sun, 

 further experiment is required for the purpose of ascertaining whether 

 it is due to radiation or to reflection. 



Note. — An opportunity occurred on the 7th of June of repeating 

 the experiments with screens at altitudes of the sun exceeding 50°. 

 The following results were obtained : — 



* In the above experiments, it is evident that the whole of the results were not 

 due to direct radiation or reflection. Account must be taken of the greater or 

 less distance of the heated surface of the ground, and of the hot air in contact 

 with it, from the bulb of the solar thermometer. 



t Experiments were also tried with a 7-inch lens, without result. 



% The thermometer exposed to the sky would probably have stood lower 

 than the one in the house if the sky had been perfectly clear. 



