358 C. 0. Ilutchins — Radiation of Atmospheric Air. 



many accidents the bars in use are made a little stouter, about 

 0-008 cm thick and 0'05 cm wide, and are united to a very thin 

 disc of copper foil - l cm to 2 cm in diameter. 



Such a junction has, when the component bars are each 

 O5 0m long, a resistance of four or five ohms, and consequently 

 the galvanometer has been re-wound to about the same resist- 

 ance. The result is an improvement in the sensitiveness and 

 steadiness of the apparatus. 



On account of the peculiar conditions under which we must 

 work in determining the radiation of air we seem practically 

 confined to Tyndail's method, using a moving mass of air, inas- 

 much as gases are nearly transparent to the rays from all solid 

 bodies at ordinary temperatures. We can however improve 

 the method in detail, making use of a moving column of air, 

 hotter, but at the same pressure as the surrounding atmosphere, 

 and moving at such a rate through it as to preserve definite and 

 measurable dimensions. The following experiments deal en- 

 tirely with air in the ordinary condition, neither purified nor 

 dried. Little would probably be gained by using purified and 

 dried air, for, as will be seen later, the radiation is mainly from 

 the surface of contact between the hot and cold air, where 

 more or less mixing must take place, so that the loss of heat 

 from contact of hot and pure air, and cold and impure air, 

 would in most respects be a less definite problem than that of 

 the radiation from the ordinary air of homogeneous even if of 

 somewhat variable and uncertain composition. 



The variation in composition causes great and rapid altera- 

 tions in atmospheric radiation, and as we cannot allow for the 

 variations the only remedy seems to be to work rapidly when 

 the conditions are favorable. 



It has in fact been learned, at the cost of much time, that 

 nothing can be gained by working under any but the most 

 propitious circumstances. We must have nearly constant 

 temperature and humidity, absence of wind, etc., if we are to 

 expect consistency in observations made at even as short an 

 interval 'as an hour. However, we have forcibly brought to 

 mind the remarkable power that slight variations in the 

 ordinary composition of the atmosphere have in altering its 

 heat-radiating properties. 



Apparatus. 



A galvanized iron pipe three feet long, two and one-half 

 inches in diameter is supported at an angle of about forty-five 

 degrees and heated by one or more Bunsen lamps placed 

 underneath. A large hole is made near the bottom of the pipe 



