I>edii8lons refpeBitig the Power ofWtndmlllSf ^c. 46« 



But to return to our fubjecl. In this praftical and popular communication I fliall avoid 

 entering into any difcuffion of the theory of windmills, which has employed the attention 

 of fo many eminent men for a century paft. The advantages of thefe engines for fuch 

 work as will admit of being performed and fufpended by intervals are fuflicientiy known, 

 but perhaps not fulEciently attended to. There can be no doubt but that the Dutch, who 

 ufe windmills for fawing, pumping, arid various other works, as well as grinding, muft 

 have found them very profitable, fince their country Is overfpread with them. It may be 

 proper however to take notice, that many writers have copied one from another the deter- 

 mination of Maclaurin, that the bed angle for windmill-fails to make with the line of di- 

 re£lion of the wind is 54° 44', which is only true, as that excellent mathematician obfervcs, 

 at the very commencement of the motion, and requires to be enlarged as the velocity of 

 the fails increafes : for the law of which, and other eflential obje£ls, his account of Sir 

 Ifaac Newton's Philofophical Difcoveries, and his Treatife on Fluxions, may be confulted» 

 Smeaton, who had much experience in the bufinefs of a civil engineer, and whofe data 

 may always be depended on, though his theories are not conftantly accurate, made a fet 

 of experiments on the conftru6lion and efFefts of windmill-fails, which are defcribed in 

 the Philofophical Tranfaclions for the year 1759*. This engineer ftates, that the mean 

 rate of work for mills with Dutch fails is when they make about thirteen turns in a minute, 

 which is when the velocity of the wind is 8y miles in an hour, or ii\ feet in a fecond : 

 and this wind in common phrafe would be called a frefh gale. Taking the maximum of 

 Defaguliers, hereafter to be mentioned, as his ftandard for computation, he deduces the 

 fize of a windmill-fail of the figure juft mentioned, and alfo according to a figure con- 

 ftru£led from his own experiments, which Ihall be equal in mean power to one man ; and 

 tlience he arrives at the inference, that one of his own fails, thirty feet in length, will, 

 when working at a mean rate, be equal to the power of 18,3 men. Ke had an oppor- 

 tunity of verifying this in the large way in a mill ufed for cru filing rape-feed. The mean 

 power of a windmill is therefore very confiderable j but what may be the annual or ave- 

 rage quantity of work fuch an apparatus is capable of performing under all the viciflitudes 

 of the wind, I poffefs no means of afcertaining. 



In the fame treatife Smeaton makes feveral very jufl: remarks on thofe windmills which 

 are a£led upon by the dire£t impulfe of the wind againfl: fails fixed to a vertical fliaft. His 

 objections have, I believe, in every infi:ancc been juftified by the inferior eflicacy of thefe 

 mills when compared with the charges of ere£ling them. He alfo maintains that water- 

 mills with oblique fails, upon the principle of the common windmill, cannot prove benefi- 

 cial to the undertaker. It is indeed probable that mod of the circumftances of running- 

 water are likely to render the common over and underfliot-wheels cheaper and more ef- 

 feftual, and that the oblique float-board will in no inftance come near the efi^ecl of a clofe 

 overfliot wheel. But it is alfo certain that fuch wheels are ufed in China, in the fouth of 

 France, and elfewhere, with much more tSodc than Smeaton- appears difpofed to think 

 them capable of : and at all events, the fubjefl of thefe wheels deferves to be confidered. 



'* This account has fince been repiibilihed, together with o\\\t\i valuable papers of the fame author, under 

 the title of kn Experimental Inquiry concerning the Natural Powers of Wind and Water to turn Mills, &c. 

 By John Smeaton, T. R. S. 8TO,--printed for Taylor, London, 1794. 



-Vol.. n.— Jan, -1 799'. 3O ^ The 



