500 Mr. J. J. Waterston on Solar Radiation. 



a sensible difference was caused by the varying amount of the 

 stem that was under the influence of the sun's rays as it moved. 



§ 5. The observations taken on the morning of the 21st of 

 August, continuously during 2-Jr hours of uninterrupted sunshine, 

 were graphically equalized, the curve drawn, and ordinates mea- 

 sured off at every 20 minutes. This was the only opportunity 

 that occurred of continuous observation between such favourable 

 limits of altitude as to indicate the direction of a line with some 

 precision. 



§ 6. In the Table of observations given in the Appendix*, the 

 date and apparent time are given in the first two columns. The 

 timepiece was regulated daily by the one o'clock signal-gun. The 

 third column contains the values of r, the observed difference 

 between thermometers X and Y. The film of talc that was 

 interposed between the sun and X was found to reflect j%ths of 

 the incident rays. This ratio was determined by observations 

 taken with the film off during calm weather. The value of r 

 without the film to /• with the film on was as 1*18 to 1*00, the 

 sun's power not sensibly varying during the interval. This pro- 

 portion was maintained at low values of r, and even when the 

 source of radiation was a gas-flame. The fourth column contains 

 the observed values of r increased in this ratio. The fifth 

 column contains the corrections required to reduce the values in 

 the third to a vacuum. The correction is taken from a scale 

 that was constructed by means of an empirical formula derived 

 from observations on the cooling of X, as detailed in the Appen- 

 dix. The sixth column is the final value of r as it would appear 

 in a naked vacuum, that is, a vacuum without any interposed 

 transparent solid between the sun and the bulb of the thermo- 

 meter. The numbers in this column represent the quantity of 

 heat -force supplied from the sun to the bulb of the thermometer 

 in a constant element of time, or the quantity that emanates from 

 the bulb in a unit of time. 



§ 7. The experiments on the cooling of the thermometer X 

 in a vacuum show that, from r = oQ z to r=15°, the time of cool- 

 ing was 294' beats of a time-piece, of which 77J were equal to 

 60 seconds; also from r = 15 c to r = 7^° the time was the same, 

 and generally from r = 2m to r—m the elapsed time is the con- 

 stant 294, which thus represents the logarithm of 2 in the loga- 

 rithmic curve of which the ordinates are r, and the abscissae the 

 time of cooling, t. The equation of the curve being 



clog- 1 =# 1 — 1 { 



OJ 



* The observations commenced on July 28, and continued for one 

 month, at times when the sky was sufficiently free of clouds. A selection 

 from this Table is given at the end of this paper. 



