Internal Work of the Wind. 433 



A second instrument was a very light anemometer, having 

 paper cups, of standard pattern and diameter, the weight of 

 arms and cups being only 74 grams, and its moment of inertia 

 8604 gr. cm. 2 



With this instrument a number of observations were taken, 

 when it was lost by being blown away in a gale. It was 

 succeeded in its use by one of my own construction, which 

 was considerably lighter. This was also blown away. I 

 afterwards employed one of the same size as the standard 

 pattern, weighing 48 grams, having a moment of inertia of 

 11,940 gr. cm. 2 ; and, finally, I constructed one of half the 

 diameter of the standard pattern, employing cones instead 

 of hemispheres, weighing 5 grams, and having a moment of 

 inertia of but 300 gr. cm. 2 



In the especially light instruments the electric record was 

 made at every half revolution, on an ordinary astronomical 

 chronograph, placed upon the floor of the tower, connected 

 with the anemometer by an electric circuit. Observations 

 were made on January 14, 1893, during a light wind having 

 a velocity of from 9 to 17 miles an hour ; on January 25 and 

 26 during a moderate wind, having a velocity of from 16 to 

 28 miles an hour ; and on February 4 and 7, during a 

 moderate and high wind ranging from 14 to 36 miles an 

 hour. Portions of these observations are given on Plates VI., 

 VII., and VIII. A short portion of the record obtained with 

 the standard Weather Bureau anemometer during a high 

 north-west wind is given on Plate IX. 



A prominent feature presented by these diagrams is that 

 the higher the absolute velocity of the wind, the greater the 

 relative fluctuations which occur in it. In a high wind the 

 air moves in a tumultuous mass, the velocity being at one 

 moment perhaps 40 miles an hour, then diminishing to an 

 almost instantaneous calm, and then resuming *. 



The fact that an absolute local calm can momentarily occur 

 during the prevalence of a high wind was vividly impressed 

 upon me during the observations of February 4, when, 

 chancing to look up to the light anemometer, which was 

 revolving so rapidly that the cups were not separately dis- 

 tinguishable, I saw them completely stop for an instant, and 

 then resume their previous high speed of rotation, the whole 

 within the fraction of a second. This confirmed the suspicion 

 that the chronographic record, even of a specially light ane- 

 mometer, at most but imperfectly notes the sharpness of these 



* An example of a very rapid change may be seen on Plate VIII. at 

 12,23 p.m. 



