432 Proceedings of the Eoyal Irish Academy, 



Tkircily, the centrifugal force due to tlie circular track of tlie 

 Anemometer's axle must make it press against its upper bearing, and 

 thus increase the friction, unless, as in my original instrument, there 

 are mechanical antifriction appliances at this part ; but this does not 

 seem to have been the case in those used by IT. Dohrandt. "With 

 his machine, when V = 40 JT, this lateral pressure Tvdll be 3'78 times 

 the weight of the moving parts, and the additional friction at the 

 bearing about O'l of this."^" 



It is not quite correct to assume that the mean of the results ob- 

 tained by rotating the machine in opposite dii'ections is identical with 

 what would have been given by a rectilinear motion ; for it cannot be 

 doubted that the equation which gives the relation between V and v, 

 must contain F-, v''-, and Vv ; and it is obvious that half the sums of 

 these will ditPer from those of the mean, F'and v, though not consider- 

 ably. M. Dohrandt tried to avoid the necessity of this double rota- 

 tion by (as I had done) making the plane of the Anemometer's rotation 

 perpendicular to that of the arms. In this case the direction of the 

 rotation should make no difference, and the centrifugal cui'rent being 

 perpendicular to the Anemometer plane should, but for the eddies, 

 have no effect. f But this is far from being the case. Allowing O'Oo V, 



V . 

 for the circular current, the ratio _ is in the two cases 



V 



V V 



V- 21= 32-01 K. — = 3-050 ; V- 21= 32-37 K. —2-770 



V V 



,r . 24-57 2-558 22- U 2-809 



°" 19-53 3-179 21-84 2-805 



18-74 3-362 15-13 2-974 



He remarks that nothing can be made of this, and the reason is ob- 

 ■vious, for the axis of the Anemometer is only 2-8 feet from the roof 

 and its cups 2-3 feet. In a fi'eer space, he remarks, this mode would 

 be preferable to the one he employed. "With respect to it, I think the 

 preceding remarks will suffice to show that it is very doubtful whether 

 his V represents the wind which it is supposed to represent, and that 

 instead of deducting any allowance for the circular current, it might 

 be nearer the trathto add one for the centrifugal forces and the eddies; 

 and certainly the resistance is different from that which would occur 

 in the ordinary use of an Anemometer. 



The railway experiment is not more conclusive. Xo reliance can 



* In one of Casella's construction, of the same dimensions as C 318, the moveable 

 parts weigli 3500 grs. Hence, for F=2o miles, the additional fiiction would be 

 19*6 grs. Six times the normal one. 



t An Anemometer, with 3-inch cups, fixed to the axle of the vane of my Ane- 

 mometer, so that its axis was always in the direction of the wind, made 48 revolu- 

 tions in 6™. In M. Dohrandt's experiment the eddies were probably much stronger 

 .than in the above case. 



