■t'jS On thd true Porm of the 



To this arm is fixed the ftraightwires i, 2, 3, 4, 5» ^j 7> which 

 are intended to reprcfent the wind, moving parallel, to ftrike 

 the vane*. To prove the truth of the principle, lift the arm 

 tip (as in fig. I,) till the points of the wires are as high as 

 the upper edge of the vanes, and bring the upper edge of 

 either vane immediately under the points ; then, by g*ently 

 prefling the arm down, (as fliewn hy fig. 2,) it will be cb- 

 fen'ed that the vane is turned by all the points preliingupon 

 it at the fame time, till they pafs beyond its lower edge. 

 Now, were the vanes any other than fpirals, this could ilot be 

 the cafe ; for, bring the wires to the plane vane AH (fig. 3), 

 (which may be fet to any angle), and by preffingthe arm down 

 as before, it will be obfcrved that, as the vane turns round, 

 tlie points, beginning from the extremity^ are quitted by the 

 vane, which at length is afted upon by the point next the 

 centre only: therefore it may be concluded, that the wind 

 would n^it continue to aft umjormly on any fly or vane which 

 is not conftrufted on the principles of a fpiral. 



The fpiral vane I, fig. i, ends, as do the reft, where the 

 anole of weather (or the angle which the vane makes with 

 the plane of its circular motion) arrives at 35", as this ad- 

 mits of the extremity of the vane being ftruck by the wind 

 with an angle of 54" ; which angle, it is demonftrated^ 

 ■will produce the greateft lateral prefTure. The (hortening 

 of the vanes of windmills is an objedi of material moment; 

 vet I do not mean to aflert that the angle of 35 degrees is 

 the beft concluding point, fince, if more power be wanted, 

 the lenoth of the vanes mud be extended. The wood, of 

 which I formed my model, is preferved at the back of the 

 vanes, (fee the dotted lines fig- 1-) merely to give a better idea 

 of their weather as they recede from the centre. Though 

 the number of fpiral vanes iliewn are fix, yet I do not fay 

 that fix are fuperior to any other number ; it being prefumed 

 that more vanes on this principle might be placed with ad- 

 van ta^-e, provided the whole area of the circle be not too much 

 crowded, which would prevent the free efcape of the wind. 



The 



