rCO OBSERVATIONS ON THE 



and fancy in place of philofophy, would be fubje8:Ing 

 ourfelves to innumerable errors. Hence it is, that in 

 the conftruction of various compound machines, fuch 

 frequent blunders are committed j for inftance, where 

 it is required to find the juft proportion and dimenfion 

 of the diiferent partsota machine, for raifmg water, &c. 

 and the ratio of the weight to the power, fo that the 

 mod work might be done in the leall time poflible, 

 dimenlions are generally taken from rules which apply 

 only to a ftate of equilibrium, without having any re- 

 courfe to velocity. And hence alfo, the miftaken prac- 

 tice of loading an arch, of whatever figure it may be, 

 with the fame mafs of mafonry, without knowing the 

 principle of equilibration, whereby theextrado of any 

 arch is fo conftrufted, that every part of the aich (hall 

 fuftain a predure, juft fufRcient to retain it in its perfeft 

 form. But to enumerate the inftances where theory is 

 requifite, would be endlefs, becaufe it would be diffi- 

 cult to mention one (ingle cafe where it was not ne- 

 ceiTarv. The great objett to be attended to in found- 

 ing a correct theory is, to include in the data every 

 circumftance that can occur 3 and it is from this neg- 

 lect, that in mixed mathematics, authors have fome- 

 tinies differed in their rcfults, though their mode of 

 reafoning has been ftrittiy mathematical. It is to be 

 regretted that men of abilities have not paid more at- 

 tention to experimental knowledge, where they have 

 been in fearch of data for applying abftra^l reafoning 

 to the rude operations of matter. Mr. Vince, one of 

 tlie firft mathematicians of the prefent age, is now open- 

 ing a new path to the moil valuable difcoveries, by the 

 belt conducted experiments that have yet been com- 

 municated to the world, relpecting friction, and the 

 refiftance of (iuids. By the former he has difcovered 

 vcrv different laws to what have been followed hi- 

 therto, and v/hich, when conlidcred and applied to 

 compound machines, whofe clfects after being put in 

 ujotion are invefiigarcd, will tend greatly to compleat 

 trie fcience of metiianics. flis experiments lalt men- 

 tioned 



