438 Proceedings of the Royal Irish Academy. 



This decreases as v increases, and vanislies when it is indefinitely 

 increased. 



Then comes the question, how are these co-efficients to be deter- 

 mined ? Not, I fear, by any observations with actual wind, for none 

 of the methods which have been proposed to measure its velocity are 

 satisfactory. I should prefer determining them by immersing the 

 Anemometer in a stream of water, were it certain that the elasticity of 

 the air makes no diiference. But here also the velocity of the current 

 varies in different parts* of its section, and should be meaned through 

 that part occupied by the instrument. Such experiments would be 

 very desirable, when fit opportunity could be obtained ; but this is not 

 easily found, and we should, in the first instance, try to get from 

 the whirling machine its lest /possible results, which, I think, with 

 proper precautions, will be far better than those which M. Dohrandt 

 was able to obtain under the conditions of his experiments. I will 

 therefore point out, first, the mode in which it and the Anemometer 

 connected with it should be constructed ; and secondly, the way in 

 which I think it should be employed. 



To begin with the whirling machine : — It must combine a strong 

 framing with facility for transport, since it should be used in a room 

 of large dimensions ; and such are generally employed for public ob- 

 jects, from which they cannot be long diverted. The framing should 

 therefore be easily taken asunder. I think it essential that the cups 

 in their rotation should not be nearer than ten feet to ceiling, walls, or 

 floor. This, if the arms of the Anemometer be two feet, will require 

 the horizontal arm to be twelve feet from the ground. Its length 

 should be as great as the locality permits, in accordance with the above 

 conditions. If too short, there would be a difference of pressure on the 

 inner and outer sides of the cups, which might disturb the results. I 

 think eight feet would be sufficient. A single arm with a counter- 

 poise alone is required. The arm (and all that it carries) should pre- 

 sent as little resistance to the air as possible. That which I used was 

 made of sheet iron \ inch thick, filed to sharp edges, and made inflexible 

 by ties of steel wire attached to the top of the axle. It was only two 

 feet long ; but the same construction will be fully available in the 

 present instance. The vertical axle is a tube strong enough to resist 

 the chiving force ; it turns in collars, one at the top of the frame, 

 another carried by cross-pieces about two feet from the ground. This 

 last has a disc, on which, by means of three balls, or by conical rollers, 

 and a flange attached to the tube, this latter revolves. This perfora- 

 tion of the axle — an idea for which I am indebted to Mr. Grubb — 

 simplifies greatly the mechanism of the apparatus. 



A cord passing to the brake-lever of the Anemometer over a pulley 

 at the top of the tube-axle supports an inner tube, adjusted central to 

 it by guides. This tube carries, below the axle, a stage on which cir- 



* A? to this see page 430, note. 



