The Cooling and Evaporative Powers of the Atmosphere. 439 



the gauge and tunnel and determined by comparison with a Pitot tube, and 

 d the difference of level of the liquid in the two arms of the gauge. 



In all the observations taken the kata was heated in hot water in a 

 Thermos flask, being kept in the water till the air space at the top was about 

 half full of alcohol. The same kata, stop-watch, and thermometer, the last 

 graduated to 0-1° C, were used throughout the experiments. In the Eoyal 

 Society paper cited above it is detailed how a factor is determined for each 

 kata by which the readings can be expressed as cooling powers in mille- 

 calories per square centimetre per second. 



The value for the kata factor was re-determined before being used in an 

 enclosure of still air jacketed with water, the results obtained being as 

 follows : — 



Mean time of cooling = 105"9 seconds. 

 Temperature of enclosure = 17"8° C. 

 e = 36-5°- 17-8° = 18-7° C. 



The factor = 0-27^ x mean time of cooling (105'9) = 535. The graph 

 obtained from the result of these later experiments and plotted in a similar 

 manner to that obtained from the earlier ones, gives the value of the constant 

 a in the formula H/^ = 0"27-t-a\/V' as 0'49. The figures obtained are given 

 in Table 2. 



The difference in the value of the constant found from the two sets of 

 experiments suggested that the true mean velocity had not been obtained in 

 the earlier work, and on re-considering the problem it appeared probable 

 that if half the area of cross-section of the kata in place of the whole were 

 subtracted from the sectional area of the tube used in the earlier work a 

 more correct value of the velocity would be given. Calculating the velocity 

 in this way the value of the constant becomes 0-36 x 1-32 = 0'48, which 

 is in close agreement with the value obtained from the later experiments. 



We tested the new constant by using it in measuring the velocity of the 

 wind with the kata against the standard anemometers at Kew Observatory, 

 and we are much indebted to Dr. Chree for the facilities lie gave us to 

 do this. 



A succession of kata readings were taken during each period of time and 

 the mean of these taken, while the anemometer records were continuous. 

 An extra strong gust may thus be missed by the kata, if no reading happened 

 to be taken at that moment. This possibly was the case in the third series 

 of observations given in Table 1. 



Table 1 shows the values obtained for wind velocities by a Eobinson's 

 cup anemometer, a Dines pressure tube, and a kata-thermometer. The 



