MADE WITH A NEW ANEMOMETER. 307 



Several questions obviously offer themselves respecting the numbers 

 thus registered. What is their real import? How far is each instru- 

 ment consistent with itself? In what manner are two such instruments 

 comparable ? 



I cannot at present answer these questions completely, but I will 

 make a few observations on each. 



As to the import of the indications of this Anemometer, it is evident, 

 that their magnitude will increase with the force of the wind, and with 

 the time to which each number refers. If we could assume that the 

 velocity of revolution of tlie fly of the Anemometer is always propor- 

 tional to the velocity of the wind, the space which the index passes over 

 on tlie scale would be proportional to the velocity of the wind, and the 

 time during which it has blown, jointly ; that is, to the total quantity 

 of the aerial current which has passed the point : and however the 

 velocity of the wind might vary, the instrument would give the sum 

 of all the elements of the current, or in other words, would integrate 

 the velocity multiplied into the differential of the time. Hence I 

 term the amount registered by this instrument the Integral Effect of 

 the wind. Tliat the velocity of the fly is thus proportional to that of 

 the wind, I have not yet ascertained ; and till that is done, I can only 

 urge, that it appears highly probable that the instrument will afford at 

 least some approximation to such a result; which no instrument hitherto 

 erected, so far as I am aware, has ever pretended to do. 



The question whether the instrument be consistent with itself, is 

 one of considerable difficulty ; for it does not readily appear how we 

 are to obtain any permanent standard by which we may test its indica- 

 tions at different times, and thus ascertain whether its scale has varied. 

 It is certainly very conceivable that the friction and other impediments 

 to motion should alter considerably from month to month, so as to 

 affect materially the rate at which the instrument would move with a 

 given wind. We might however imagine means by which tlie actual 

 velocity of the current of air which turns the instrument should be 

 ascertained, and thus this difficulty overcome. For example, the Ane- 

 mometer might be placed on some part of a large machine which mov^s 



Vol. VI. Part II. II u 



