366 Eicaminatkn of a new Pijott •with metallic Fittings. 



fame curvatufe. If his pifton be prefled into a fmaller cylinder than correfponds with the 

 a£lual circle his pieces are adapted to form, thofe pieces will be prefled in, and will touch 

 only at their extremities ; fo that every joint will be immediately over a place where the 

 oppofite ring does not in f aft touch the cylinder : it will therefore,' in this cafe, leak prin- 

 cipally at the joints. And, on the contrary, if his pifton be prefled into a larger cylinder 

 than correfponds with the circle of the rings, each portion of thofe rings will touch the 

 cylinder in one point only. The moft favourable point will be at the half-way between the 

 extremities of each piece. In this pofition the joints indeed will not leak, but every other 

 part of the circumference will; and the places of the moft open paflage will be at thofe 

 points of the circumference which are equidlftant between joint and joint. Thefe confide- 

 rations relate to perfect circles ; but if we attend to fmaller irregularities, whether convex- 

 ities or concavities in the cylinder, it appears evident, that the fegments of the rings being 

 inflexible will ftill lefs efFe£tually adapt themfelvesto fuch imperfections. 



Thefe objeftions, relating to the figure of the periphery of this metallic pifton, are of 

 very ferious import : thofe which relate to its action are fcarcely lefs fo. The furfaces of 

 the rings thus laid upon the pifton and upon each other, together with the furface of the 

 piece which confines them from rlfing, muft all be very well adapted to each other ; and 

 the number of fquare inches of this furface muft, in the nature of the contrivance, be con- 

 fiderable enough to afford much fri£tion. It may reafonably be doubted whether thefe 

 pieces, in the rapidity of afcent and defcent, can obey the a£tion of the fprings during the 

 very fhort times in which the pieces are oppofed to the irregularities they are meant to 

 remedy. But when a great preflure, fuch, for example, as the readlion of a column of- 

 100 feet of water, comes to be exerted upon the face of this apparatus, the plates or pieces 

 of thefe rings may be imagined to be confined in a vice. The preflure of fuch a column will- 

 amount to more than 40 pounds upon every fquare inch. Whence we may conclude, 

 ieither that they would not move at all, or that the force of the fprings muft be fuch as 

 greatly to load the work with fri£tion, and damage the apparatus by fpeedy wear. 



Laftly, it feems to be a queftion or doubt, which well deferves to be refolved by further 

 experiment, whether, in any cafe of reiterated or long continued a£tion, the fofter metals' 

 can be made to work in contaft with each other, in the way of clofc fitting, with as little 

 tefiftance and wear as when an organifed fubftance containing oil or fat is interpofed. 



When one individual fpeaks to the world concerning the works of another, the tranf- 

 aftion is naturally accompanied with a fenfe of perfonality. As this fenfe ought not to 

 kad men into unworthy a<£tions, fo, on the other hand, it ought not to prevent their ful- 

 filling any duty which may call upon them. Convinced that no vindication or apology is, 

 neceflary in defence of a fcientific examination of every objedl which is offered to the- 

 public acceptance, I have fpoken freely of a conftru£lion which, from its ingenuity, might; 

 be thought of greater value than it really is ; but which, when carefully examined, appears, 

 to be Inferior to the methods already in ufe. 



xir //;- 



