the Robinson Cup -Anemometer. 77 



our further investigations which is the correct sign, so that 

 there is no necessity to come to a final decision. 



The Steady State. 



§ 10. The first application to be made of (14) is to the 

 steady state. 



In high winds, as already indicated, a close approximation 

 to the value of Y/v is supplied by (19). Now for the same 

 anemometer, at least when in a uniform state of lubrication, 

 a 2y f>2) c 2 are constants, so that in high steady winds Y/v is 

 practically independent of V. 



Each of the coefficients a 2 , b 2 , c 2 contains a part arising 

 from the wind's driving action which vnries with the arm Z, 

 and a second part arising from the friction which is associated 

 with the horizontal pressure produced by the wind's action ; 

 the latter part is practically independent of l, but varies with 

 the coefficient of friction, and with the arm d at which the 

 frictional forces act. Thus, if two anemometers in which 

 the friction and djl are widely different give nearly identical 

 values of Y/v in high steady winds, then there is a strong 

 presumption that the frictional forces of the species considered 

 are unimportant in both. The ratios a 2 : b 2 : c 2 do not depend 

 directly on the weight or inertia of the instrument, so in 

 high steady winds these factors would seem of secondary 

 importance. 



In light winds — a — a x v — b^V may no longer be neglected. 

 The coefficients a 1? ^ depend presumably rather on the peri- 

 meter than the area of the cups, while a depends on the 

 weight and the lubricant. There is thus no apparent reason 

 for expecting in light winds any close agreement between 

 instruments of different patterns variously lubricated. As 

 a matter of fact, in very light winds one anemometer may 

 stand still while another, precisely similar in outward appear- 

 ance, a few feet distant goes on recording a mile or two an 

 hour. 



Some idea of differences in this respect may be derived 

 from the consideration of the wind-velocity V which just 

 suffices to keep the cups turning. This is the positive root 

 of (15), or 



y o _ K + vV + 4oo£2 # (20) 



%c 2 2c 2 



Other things being similar, V increases with a , and so with 

 the weight of the apparatus and the coefficient of friction. 

 The less limpid the lubricant the larger V . Thus in light 

 winds the readings even of one and the same anemometer 



