RAILWAY CONSTANTS. 199 



tried, the index oscillates between zero and the extreme limit of 

 its play. 



Besides this difficulty, which, from the nature of the resist- 

 ance, would appear to be inseparable from every form of dyna- 

 mometer, another will arise if it be admitted that the atmo- 

 sphere have any considerable share in producing the resistance 

 which the tractive power has to overcome. The dynamometer 

 must be interposed between the engine and tender, or between 

 the latter and the first coach or wagon in the train ; or, to speak 

 more generally, it must be immediately before the coach or 

 wagon whose resistance it is used to measure, and must be 

 behind the engine, tender, or carriage which precedes that load. 

 It is evident that, under such circumstances, the atmospheric 

 resistance will produce only a modified and partial effect on the 

 dynamometer; nor will this instrument, under such circum- 

 stances, exhibit a true estimate of the resistance arising from 

 friction alone, independently of the atmosphere, since the effect 

 of the atmosphere is only partially intercepted by the preceding 

 part of the train. 



The only manner in which the dynamometer could be used 

 with any prospect of obtaining a tolerably correct and satisfac- 

 tory result, would be to construct it in such a manner as to re- 

 gister its own indications, by describing a curve on paper with 

 a pencil moved by the index of the instrument, so that the ordi- 

 nate of this curve would represent the resistance, and the cor- 

 responding abscissa the point of the road where that resistance 

 was produced. If the instrument thus constructed were applied 

 with so very slow a motion as to render the atmospheric resist- 

 ance so small that it might be practically disregarded, then the 

 mean value of the ordinate of the curve, or, what is the same, 

 the area of the curve divided by its abscissa, would express the 

 mean amount of the resistance. 



The second and third methods of experimenting are those 

 which have been hitherto generally used for the practical deter- 

 mination of the resistance of railway trains to the tractive power. 

 The combination of both has been resorted to by M. de Pam- 

 bour in the following manner : — He placed the cai-riages whose 

 resistance was to be determined upon a steep inclined plane, 

 having a line nearly level at its foot, and allowing them to move 

 by gravitation from a state of rest, they attained a certain velocity 

 at the foot of the plane ; with this velocity the carriages moved 

 along the level until they were reduced to a state of rest. It 

 was then assumed that the resistance was represented by the 

 ratio of the difference of absolute levels of the point from which 

 they started and the point at which they stopped, to the di- 



o 2 



