378 F. W. Very — Sky Radiation and Isothermal Layer. 



appeared presently, for at 6 h 30 m a. m., over an hour later, the 

 air temperature by a fan-ventilated instrument one meter above 

 the surface was -i-3'0° F., dew-point = +2*0°, relative humid- 

 ity=0*94, while the snow remained at —6*0° F., or the 

 depression had increased to —9°. In still another hour, at 

 7 h 35 m a. m., the sun had risen, the sky remained as before, the 

 air temperature was +2*2° F., relative humidity =100 per 

 cent, but the shaded snow on the west side of the building, 

 and about 5 meters from it and freely exposed to the sky, 

 remained as before at precisely — 6*0° F. The air was per- 

 fectly calm. A dense fog had formed in the Neponset valley, 

 but did not reach our level. Soon after this, a gentle breeze 

 sprang up and the snow rapidly assumed the same temperature 

 as the air. 



It is evident that during the whole of the earlier observa- 

 tions, a thin layer of supersaturated air adhered to the surface 

 of the snow. The radiant absorption by air supersaturated 

 with moisture is extraordinarily great, and, in this case, it pre- 

 vented the snow from receiving radiation, while downward 

 conduction from the slightly warmer air at one meter was insig- 

 nificant, and thus the snow retained the temperature which 

 had been given to it by nocturnal radiation to a clear sky 

 before the cloud appeared. A similar diagnosis would, no 

 doubt, dispel many illusions in regard to records of exceptional 

 temperature depressions produced by nocturnal radiation. 



The area of sky included within the instrumental aperture 

 was about 24 square degrees, or a circle ten times the diameter 

 of the moon. The heat-measuring instrument was standardized 

 as follows : 



(1) For a temperature 33° F. below that of the instrument, 

 the radiation to a tin plate screen in a distant black shelter at 

 the temperature of the outside air (tin plate reflecting to the 

 black walls of the shelter or to the ground at air temperature) 

 gave an average galvanometer deflection of —0*64 div. per 

 degree Fahrenheit in fair weather, and —0*61 in foggy 

 weather. 



(2) Black card at outside air temperature, suspended outside 

 the aperture in the north wall of the observatory and exposed 

 as in all other cases by withdrawing a screen of tinned wood 

 within the building, having both the inside temperature and 

 very feeble radiating power, gave for an external temperature 

 24° below that of the instrument, —0*65 div. per degree F. 



(3) Snow at an average temperature 44° below that of the 

 instrument gave — 0"65 div. per degree F. 



(4) A black radiator (boiling and near at hand) having the 

 cavernous shape appropriate to an ideal radiator, and compared 

 with a screen at air temperature, gave for an excess of +192° F. 



