PARALLEL EULE3. 197 



play in the Aides, or deviation in the grooves in which ihey becomes inaccu- 

 are to move, muft alter the parallelifm of the lines drawn by "^^ from wear. 

 it; and however exadt it may be at firft, the natural wear at- 

 tendant on its ufe muft demonftrably produce thefe imperfec- 

 tions; to which may be added, (hat the nicely of workman- 

 fliip which it requires, and its complicated form, muft of courfe 

 render it expenfive. 



The inflrument formed by a rule fupported by two Imall That rriovlng 



wheels fixed to the fame axis, which axis is placed fo as lo^^"?.'^*'/^ ^' ^ 

 ' 1 fixed to one axis, 



be parallel to the edge of the rule, is liable to be imperfedt, has the fame 

 from any difference in the diameters of the two wheels, or f.'^f^'^*' ^"'^ '^ 



•' liable to inaccii- 



flight inaccuracy in the pofilion of the axis. racy from the 



This rule is alfo very liable to flip on the paper, and js ""^^'^""'=*s of 

 rendered incorrect in its effects by any unevennefs in the fur- 

 face over which it is moved. 



The parallel rule, mentioned in your ninth volume, page That formed by 

 '212, requires an exact proportion in the length of each of its ^ob^diffiJulTto 

 parts; and as thefe are all of ditlerent meafures, would be make exad, and 

 liable to error in the firft formation, on this account ; and how- ''^ble to become 



^i , 11- II r 1 • '^^'■y 'ncorreft 



ever exactl\ made, would, aher a hllie wear, foon deviate, in wear from its 

 on account of the play which this would produce in the joints; '""« projeaions 

 the connedors alfo between the two rules, palling from differ- ^f fupport. 

 cnt extremities, and leaving long fpaces beyond the points of 

 fupport, would thereby occafion any play in the joints to pro- 

 duce a greater deviation Irom parallelifm in the lines drawn. 



The apparatus for producing parallel lines formed by the The drawbg 

 drawing board and normal fquare, can hardly with propriety ,^^[(-^^"^.2"°'^' 

 be clafTed among the initrumenls here treated of; w hatever its cumbrous, and 

 accuracy may be, its cumbrous form, and the time required ^^ --s time. 

 for fattening tlie paper to it, render it lor many purpofes very 

 inconvenient. 



Thefe confiderations induced me, about the time when the 

 account of the parallel rule, given in your ninth volume was 

 publiflied, to conlider how a parallel rule might be conftru6fed 

 not liable to fide deviation, and as free as poffible from the de- 

 feds of the others above ftated. The inflrument which then 

 occurred to me as the beft calculated for this purpofe, I fliall 

 now defcribe; and as I have often examined it fince, if it pof- 

 feflTed any material defed, it is probable it would have become 

 manifeft before this ; in which cale 1 Hiould not have brought 

 it forward to public notice. 



My 



