NO. 4 SOLAR RADIATION — ABBOT, FOWLE, AND ALDRICH 



39 



after the flight, against standardized pyrheliometers, and so the 

 results were reduced to calories per sq. cm. per minute. 



SOURCES OF ERROR 

 I. EFFECT OF AIR CURRENTS 



In relation to the important point raised by Mr. Very regarding 

 the effect of a downward current of air, a balloon pyrheliometer was 

 calibrated in a current of air. The method of doing this is shown in 

 figure 6, in which a b represents a 20-inch pipe connected to the 



Fig. 6. — Testing the Balloon Pyrheliometer in Air Currents. 



blower c, and causing the current of air of known velocity to pass 

 over the balloon pryheliometer d. In this situation the balloon pyr- 

 heliometer was compared, with and without flow of air, with the 

 standardized silver-disk pyrheliometer. The rate of flow of the air 

 was taken at 5 meters per second, which would be the maximum rate 

 of ascent of the balloon during its flight. 



The results of these experiments were surprising to us, for we had 

 assumed, with Mr. Very, that the effect of the downward current of 

 air would be to increase the rate of cooling of the aluminum disk 

 when the shutter was open. The contrary appears to be the case, for 



