10 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 71 



glass over the strips. This is due for the most part to wind blowing 

 on the strips. However, it is entirely possible and in fact probable 

 that during the totality there would be a change in the direction 

 found, because the atmosphere undoubtedly cooled to an appreciable 

 extent during totality. 



The values obtained during totality are about the same as those 

 obtained during the night preceding. 



The results of the preceding observations made with the pyra- 

 nometer with the glass removed and purporting to give the outflow of 

 radiation from a perfectly radiating horizontal surface in calories 

 per sq. centimeter per minute must be regarded as provisional, as 

 also all such values hitherto obtained with the pyranometer on the noc- 

 turnal radiation and on the outgoing radiation during the totality of 

 the eclipse of June, 1918. The values are computed on the assump- 

 tion that the blackened strips of the pyranometer radiated 95 per cent 

 as much as a perfect radiator of the same dimensions would do. On 

 account of the uncertainty existing as to the absorptive and radiative 

 powers of lamp black for rays of very great wave length, the value 

 95 per cent must be regarded as a mere assumption. Comparisons 

 have been made between the pyranometer and the Angstrom pyrge- 

 ometer which has been much employed for measurements of noc- 

 turnal radiation. These comparisons indicate that either the 

 pyrgeometer reads much too high or the pyranometer much too low, 

 when the pyranometer observations are reduced as above explained. 



Recently Mr. L. B. Aldrich has constructed and tried on Mt. 

 Wilson a new instrument for measuring nocturnal radiation which is 

 designed on the principle of the hollow chamber black body. Pre- 

 liminary comparisons between this instrument and the pyranometer 

 seem to show that we ought perhaps to use a coefficient of radiation 

 for the latter under 95 per cent, or more nearly 80 per cent, so that 

 perhaps the values here given should be increased by about 16 per 

 cent, and also those published by Mr. Aldrich in his paper entitled 

 " The Smithsonian Eclipse Expedition of June 8, 1918." However, 

 the matter is not yet settled and requires further Investigation. 



