942 TRANSACTIONS OF THE AMERICAN INSTITUTE. 



periods, therefore, erroneous in almost every possible manner. The 

 error was discovered before testing the second engine ; but the com- 

 petitors objected, and, as it appeared, with reason, to testing one 

 engine with any different apparatus from the other. The brake, 

 therefore, served to restrain the engine with any desired degree of 

 force, by weighting the lever, G, to the desired extent ; and this was 

 the only function performed by this portion of the apparatus. 



The perfection of tlie regulation was, in addition to the usual 

 metliod of timing by watch, tested by a slight apparatus, improvised 

 for the occasion, with very fair success. Timing in the ordinary man- 

 ner, by counting the revolutions in a minute, or quarter of a minute, 

 suffices very correctly to ascertain the average rate of a machine 

 during those periods. But no method in general use suffices to show 

 the fluctuations in speed at different periods therein. To ascertain 

 this latter point as exactly as possible, the paper-feeding apparatus of 

 a common recording telegraph instrument was employed to move a 

 slip of paper steadily forward, while a pencil, secured on the shaft of 

 the engine, made, at each revolution, a transverse mark across it. By 

 comparing the length of these graduations, an estimate was formed 

 of the uniformity of the speed. If the motion of the paper could 

 have been absolutely uniform, and the pencil secured so that no elas- 

 ticity in its fastening, or end play of the shaft, could affect its indica- 

 tions, the record thus obtained would evidently have been mathemat- 

 ically perfect. In practice, the errors were probably very slight, and 

 the results obtained may be considered quite reliable. 



The only basis on which to estimate the power of either engine, 

 was the cards or diagrams obtained by the indicator. Believing it 

 impracticable to ascertain the quantity of steam consumed in short 

 periods of a few minutes, or hours, either by measuring the water 

 evaporated in the boiler, or the fuel consumed, the only data from 

 which the quantity of steam, and consequently of fuel, used by either 

 engine, under any given circumstances, were also obtained from the 

 indicator diagrams. The uses of the indicator, and also its faults, as 

 it is commonly employed, are probably familiar to a large portion of 

 your readers. The power was deduced by measuring the area of the 

 card or diagram, and tlie quantity of steam used was obtained from 

 the terminal pressure, or the pressure of steam in the cylinder at the 

 end of the stroke. I sliould have remarked, at an earlier period, that 

 both engines were regulated by changing the point of cut-off, or, in 

 other words, by varying the degree of expansion to Avhich the steam 



