t'^aiit'S of a Vertical IVindmill. ij^ 



of turns that would be made in a given time by a ventilator 

 ki a window, the fly of a fmoke-jack, or the vanes of a ver- 

 tical windmill, with a given velocity of w ind, we nmft con- 

 fider each vane as a fpiral curve ; for fuch a vane, whea 

 llruck by a refilling fluid, muft of neceflity pafs through a 

 triangular fpaee, of which the fprcad of the extremity of tlie- 

 vane forms the bafe, in the time that the wind pafles through 

 the depth of the vane : therefore, fri(Stion and gravity being 

 out of the queition, a fiv of one vane will pafs round as quick 

 as a flv with anv greater number. 



Some years back I made a mniiber of experiments upon 

 fpiral flics moving in water, to prove what number of turns 

 the fpiral would a/luallv niake in paffing through a certain 

 fpace in the direction of its axis; and I can with confidence 

 aflert, that (if the fpiral be nicely made, and be fuch as fliall 

 be generated by a radius moving angular, and at the fame 

 time perpendicular to the plane of its angular motion*) ihe 

 revolutions it w ill make in paflling through a certain fpace 

 of water in the direction of its axis, lu'tll admit of caktdation; 

 but that, if the flies deviate from a fpiral, the revolution-J 

 arc not to be calculated iv'ilh cxaStncfs., being ever Icfs than 

 the perfect fpiral wiJl perform. "Wind is a Ifuid \\\ fonie re- 

 fpecls fimilar to water, particularly when applied to the pre- 

 fent purpofe ; therefore, vuucs wiiich are to be turned by its 

 I power, ought to he Jfiruls, for, if not, thiy il ill ever perform 



For tlie purpofe of elucidating, in the moft clear and fimpic 

 I tnanner, the truth of the principle alferted, I caufed a vane, 

 I or fly, to be made. This vane is reprefentcd by fig. i, pi. IV. 

 I The vane moves round upon the pillar AI^, which is fixed 

 ! upriglit in a ftand. Tlie part of the pillar above the fly is 

 ^ii|uare ; this carries the brafs arm CG, whicli may be moved 

 up aud down the pillar, parallel to itfelf, without revolving. 



: • 1 his conftruftion niiijlit be- othcrwiCc cxprenTed, thus : A fjiriil gc- 

 BeratL'd oy the circu!»r rrjotion of a racliils, and of a line luovio;; uf right 

 Inglcti to the plane of che kiiculur iiiobio'i. 



6 1# 



