Improved Guntcy's Scale.— Perpetual Motion, 375 



I approve of this conftruftion as fuperior to every other which has yet occurred to me, 

 not only in point of convenience, but hkewife in the probability of being better executed, 

 becaufe fmall arcs may be graduated with very great accuracy, by divifions transferred from 

 a larger original. 



The circular inftrument is a combination of the Gunter's line and the feflor, with the 

 improvements here pointed out. The property of the fe£lor may be ufeful in magnifying the 

 dilFerences of the logarithms in the upper part of the line of fines, the middle of the tan- 

 gents, or the beginning of the verfed fines. It is even poffible, as mathematicians vi-ill 

 eafily conceive, to draw fpirals, on which graduations of parts, every where equal to each 

 other, will fhew the ratios of thofc lines by means of moveable radii fimilar to thofe on 

 fuch an inftrument. 



After the publication of the account in the Tranfa£lions of the Royal Society, the men- 

 tion of circular and fpiral inftruments brought to the recolledion of the late Mr. George 

 Adams, of "Fleet-ftreet, (who had feen the Itraiglit inftrument fome years before it was com- 

 municated to the Society,) that he had a fpiral engraved on a brafs plate by his father. He 

 made me a prefent of the plate, and ftiewed me, by a manufcript, thai it was conftrufled in 

 the year 1748. The fpiral has ten turns, and its external diameter is twelve inches. It 

 contains the numbers, fines and tangents, the latter being twice repeated, and may with 

 eafe be ufed to compute to four places and an eftimate figure ; which is to the full as 

 much as could be done with a common Gunter's fcale of fixty feet in length. 



Fig. 3. Plate XVI. reprefents the line of numbers drawn according to this fyftem. The 

 fe£lor ACB is ufed to meafure the ratios. One thread muft be fet to the antecedent, and 

 the other to the confequent. If the antecedent thread be then removed without altering 

 the angle to any other antecedent, the other thread will mark a confequent at the diftance 

 of the fame number of turns of the fpiral in the fame direction. In cafe the number of 

 turns ftiould proceed without the fyftem, it will be necelTary to return and reckon 

 onward from the oppofite extremity of the fpiral to complete the number. 



IX. 



On the Mechanical PrajeBs for nfford'mg a Perpetual Motion. 



In confequence of the notice * taken of Mr. Varley's attempt to produce a perpetual 

 motion, I have been requefted by feveral coirefpondents to ftate how far the mechaaical 

 fcheme for which Dr. Conrad Shiviers took out a patent in the year 1790, for the fame 

 objeft, may be worthy of attention. I have, on that occafion, mentioned the diiiiculties 

 which have prevented any clear general demonftration of the abfurdity of ihis purfuit from 

 being produced, though it has not been difficult to ftiew the fallacy of the individual nians. 

 It docs not indeed fcem eafy to enunciate the fchcmc Itfelf. What in univerfal terms is 



* Fhilof. Journ. I. 334. 



the 



