EoBi^'soN — On tht Cap Anemometer. 431 



they would be less.--' But, in the investigations which we are consi- 

 dering, the air was far from free. The part of the Hall occupied by the 

 apparatus was 27-4 feet by 26-9 feet ; 26-3 feet high in the centre, 22-6 

 feet at the walls. The rotating arms were 20-3 feet from the ground ; 

 but the cross of the Anemometer was only 17'64 inches from the roof 

 (this was necessaiy from the natui'e of the frame supporting the vertical 

 axis, which had been constructed for another purpose), and the centres 

 of the cups were from 34 inches to 30 inches distant fi'om the main, 

 line of the walls. Part of the Hall commimicating with this nearly 

 cubical space was only 11 -7 feet high, and on the east wall was a small 

 gallery- 15 '4 feet above the ground, in which an observer was stationed. 

 It is evident from these details that the air put in motion by the 

 rapid rotation of the arms (11-25 feet long, and 2-4 inches dia- 

 meter, with braces 8-68 feet long, and 0"6 inches diameter), and by 

 the fixed parts of the Anemometer must have been thrown into irre- 

 gidar eddies, interfering with its action on the curved surfaces of the 

 cups, and that they must have been differently impelled when nearest 

 to the walls, and when opposite to the angles of the room. The 

 current measui'ed by the Voltman cannot be regarded as representing 

 that in the Anemometer track, for, besides being outside that, it 

 must be modified by the gallery and observer's body. He tried to 

 examine it by small balloons filled with coal gas, and loaded so as to 

 be in ec[uilibrio, "but they travelled in-egularly, sometimes within, 

 sometimes without the Anemometer's track, sometimes above, some- 

 times below it." 



He found, also, that the revolutions of the Anemometer tned were 

 sensibly diminished, when another one, or even the small Woltman, 

 was put on the other arm 22 feet distant. (I think some useful in- 

 formation as to the motion of the air might have been obtained by 

 placing over the Anemometer a tube as long as its diameter, perforated 

 with several holes, and discharging through these jets of smoke, as 

 was done by Dr. Ball in his experiments on vortex rings). M. Doh- 

 randt has here overlooked three disturbing elements. A part of the 

 resistance to an Anemometer is work done in throwing out from its 

 centre a quantity of air by centrifugal force ; this must be interfered 

 with by the walls, &c. Another is, that the rotation of the whirling 

 machine itseK must, by the same centrifugal force, produce an outward 

 current. He refers to this as increasing the density of the air near the 

 walls, but it must escape above and below, and the stream so pro- 

 duced must help to turn the cups. What its amount is cannot, under 

 the circumstances, be determined a priori, but it probably more than 

 counteracts the cii'cular cmTent. 



* It is, ho-vrever, to be remarked, that tlie small Anemometer mentioned in the 

 beginning of tbis paper gave almost identical results -svben canied through air. and 

 when immersed in running water. 



