NO. 29 VOLCANOES AND CLIMATE ABBOT AND FOVVLE II 



for which it was designed fairly well. It had an alt-azimuth mount- 

 ing, and measurements were made with it of the intensity of the sky 

 radiation at all parts of the sky as compared with the intensity of the 

 direct beam of the sun. In this comparison the direct solar beam was 

 reduced in intensity by means of a rotating sector diaphragm, and 

 a series resistance was inserted for diminishing the sensitiveness of 

 the galvanometer when the sun was observed. Measurements were 

 made with this bolometer on the brightness of the sky on September 

 5, 6 and 7, 19 12. 



The other apparatus used was a device originally intended for 

 measuring the nocturnal radiation, and was similar in principle to 

 the Angstrom Compensation Pyrheliometer. The two blackened 

 metal strips, instead of being side by side, were placed one above 

 and the other beneath a thin horizontal plate of brass, and were so 

 protected by strips of horn that the metal strips could not give off 

 radiation except from their outward surfaces. The radiation from 

 the outward surface of each strip was free to go in any direction 

 within a hemisphere. In the use of this instrument a blackened 

 screen was placed beneath it so that the lower strip was exchanging 

 radiation only with this screen, which subtended a whole hemisphere. 

 The upper strip was exchanging radiation with the whole sky. 



In the use of the bolometer mentioned above, the instrument was 

 protected from exchanges of radiation of great wave-length by 

 means of a plate of glass. The nocturnal radiation instrument, on 

 the other hand, had no plate of glass in front of it. Accordingly 

 while the bolometer measured only the radiation of the sky transmis- 

 sible by glass (that is to say, the radiation coming indirectly from 

 the sun), the nocturnal instrument, on the other hand, measured the 

 combined effect of rays of all wave-lengths. Accordingly a correc- 

 tion had to be applied in the latter measurements for the radiation 

 which the instrument would have sent out if the sun had been in 

 eclipse. Of course this correction could not really be measured, but 

 Mr. A. K. Angstrom had made measurements at Bassour of the 

 nocturnal radiation on the morning and evening of each of the days 

 in question and these he has kindly permitted us to use. Assuming 

 that the nocturnal radiation would have had the mean of these 

 morning and evening values at mid-day if the sun had been eclipsed, 

 we may thus estimate and correct the results for the exchange of 

 long-wave rays between the blackened strip and the sky. Thus we 

 determine the solar radiation scattered from the sky to the observer 

 as indicated by the nocturnal radiation apparatus. The experiments 



