ON A CONSTANT INDICATOR FOR STEAM-ENGINES. 321 



this piston is pierced by a tube, which passes through the bottom of the 

 cylinder, and is closed by a cock placed in such a position that it may be 

 opened and the water drawn ofF without stopping the indicator. To ensure 

 a uniform lubrication, a grease-cock is inserted in the top of the cylinder. 

 To reduce to the utmost the friction which results from the pressure of the 

 spring on the spindle of the cone, the sensibility of the cone to the action of 

 the spring is rendered as great as possible by diminishing the bearing surface 

 of the spindle, and causing the spring (which is a spiral spring inclosed in 

 a tube) to press accurately on the centre of the extremity of the spindle. 

 These precautions being taken, an exceedingly sliglit pressure of the spring is 

 found to be sufficient to produce that friction between the integrating wheel 

 and the cone which is required to drive the registering apparatus. It was in 

 the correct transmission of the motion of the cone to the integrating wheel that 

 the greatest difficulties were predicted, and that the least have been found. 



The following expedient was applied to determine the whole amount of 

 the prejudicial friction. 



The instrument being put into the same condition in every respect in which 

 it would be applied to the engine, was fixed firmly in an upright position, 

 and the bottom of the lower cylinder being taken off, a hook was attached to 

 the centre of the piston working in that cylinder. To this hook weights were 

 attached, increasing in the same order in which they were increased for the 

 separate trial of the same system of springs as was now placed in the instrument. 

 When each weight had thus been added, and the pistons brought to their cor- 

 responding positions in the cylinders and the integrating wheel to its corre- 

 sponding place upon the cone, the piston-rod, carrying with it the pistons and 

 integrating wheel, was gently depressed by the hand until the lowest position was 

 found at which it would rest, and in this position the separation of the springs 

 was accurately measured. The diflPerence between this separation and that 

 produced by loading with the same weight the same springs before they were 

 placed in the instrument, was seen to be due to two causes, first, to the weight 

 of the connected system of the pistons, piston-rod, and integrating wheel, all 

 of which weight (in this state of the equilibrium of the system bordering upon 

 motion upivards) was borne by the springs ; and secondly, it was due to the 

 friction, which in that particular position of the system' was opposed to its 

 motion upwards. Now the moveable system of the pistons, piston-rod, and 

 integrating wheel being accurately weighed, and the deflection of the springs 

 due to each pound having been before accurately determined, the deflection 

 due to the former cause was accurately known, and therefore that due to the 

 latter, i. e. to the friction. The deflection of the springs due to the friction of 

 the instrument in every position of the pistons being thus found, the mean 

 deflection due to this cause was known, and therefore the mean amount of this 

 friction. A series of experiments was thus made, with great care and accu- 

 racy, by Mr. Tininie, three of the springs being successively placed in the 

 instruments, and the weights gradually added and gradually taken offl A 

 slight variation was found in the friction in different positions of the piston, 

 its greatest amount being uniformly found in the lowest position. This dif- 

 ference may in some degree be due to a slight inaccuracy of workmanship ; 

 it is, however, no doubt, principally to be attributed to the fact, that a larger 

 portion of the circumference of the integrating wheel is brought in contact 

 with the surface of the cone in the lower than in the higher positions of the 

 former, and that by reason of the milled edge of the wheel ploughing slight 

 furrows* on the brass surface of the cone ; the friction of these two surfaces is 



* These furrows have been rendered so slight as to produce no sensible injury to the surface 

 of the cone, by increasing the sensibility of the spindle to the action of the spiral spring on 

 its extrcraitv, 



18ti. ■ y 



