42 Mr Forbes's description of a nexv Anemometer. 



ed in each proportional to the times. And for practical pur- 

 poses, it will be quite sufficient to consider merely the veloci- 

 ties corresponding to the extreme deflection included by each 

 space. Let S be the extreme deflection corresponding to the 

 first cavity/, (Fig. 2,) and N = the number of balls deposit- 

 ed in any given time. Then obviously N a V^S, or it is 

 proportional to the square of the velocity and the space con- 

 junctly. For the other circular rings, for S we must substi- 

 tute S'— S and S''— S', &c. ; S, S' and S" corresponding to 

 the successive extreme deflections. Expanding N oc V^S, it be- 

 comes NocV^ X (W-| X log.i±i) and since e = ^^ we 



haveNocV^/ ^^ x log. ^ — -f ] \ which may be thus ex- 



pressed ; 



/Nt \ 2«Z: 



N oc V-'/ — log. I 1- 1 ) M 



a complex expression, but from which might easily enough be 

 extracted the value of V in an approximate manner, and from 

 the object of the inquiry, N, in all values of S is intended to 

 be a constant quantity. In the figures these spaces have been 

 made equal, nor has the proportion of the parts been there 

 particularly attended to. The receptacles for the balls ought 

 also to be made deeper than there represented, to prevent them 

 from starting from their proper places. 



I hope I have been sufficiently clear in my descriptions, to 

 show that the instrument in its nature and applications is per- 

 fectly elementary, and requires no computation in practice ex- 

 cept counting the balls, which by their situation indicate the 

 portion of time since the last observation during which the 

 wind has blown with any force in any direction. And the only 

 adjustment it requires is to collect the pellets from the vari- 

 ous spaces, and return them to the box from which they are 

 again to fall. 



When an instrument has once been constructed on the prin- 

 ciples above laid down, nothing more is required for any period 

 during which it may be used, so long as balls of the same size 

 and density are employed,* and the simple wheel-work kept 



• It has been one object of my inquiry, whether small shot might not 

 be advantageously employed, notwithstanding the great specific gravity of 



