1907.] 0)1 Mame in Gas and Petrol Motors. 515 



lOG thermal units, when the calorimeter showed 105 thermal units 

 to be present. The method appears capable of very considerable 

 accuracy. 



This method, then, enables us at present to determine the 

 apparent specific heat, and the heat flow to the cylinder in any engine 

 under it,s working conditions. The indicator, howeyer, requires to 

 be of a yerj accurate type ; and to distinguish between a small amount 

 of continued combustion, and a yariable specific heat, it was necessary 

 to design another type of indicator which did not depend upon 

 mechanical levers for its marking upon the card. I accordingly con- 

 structed an optical indicator of a new type, which gives extremely 

 accurate results. 



Before leaving this new method of research, I would like to say 

 a few words as to some valuable work which has recently been done 

 by Professor Hopkinson, of Cambridge. Professor Hopkinson has 

 attacked the problem of heat loss incurred within a closed vessel. 

 The apparatus consists of a calorimeter capable of dividing up the 

 heat flow to the sides of a closed vessel into portions incurred in 

 minute fractions of a second. In this vessel there is a thin wood 

 backing, and a copper strip is wound close against the backing, in 

 such a way as to insulate the edges. If an electric current be passed 

 through this strip, the temperature of the strip at any moment may 

 be determined by the resistance. Assume a galvanometer to be in 

 circuit, so arranged that its readings correspond to temperature. If 

 this galvanometer be in circuit, and a gaseous explosion occurs within 

 the cylinder, one diagram can be taken to show the effects of the 

 explosion so far as pressure is concerned ; a simultaneous diagram 

 can be taken showing the temperature of the copper strip. Deter- 

 minations can be made of the heat loss to the backing, and other 

 corrections, and in this manner the exact amount of heat which has 

 left the hot gases, and passed into the walls can be determined at 

 any given moment. This arrangement promises to give important 

 information as to the rate of loss in gaseous explosions, from which 

 observations some deductions may be drawn as to specific heat, and 

 as to time of termination of combustion. 



Returning to my own experiments with the optical indicator, my 

 next sUde shows an indicator card taken with the instrument, and 

 operating to produce the new alternate compression and expansion 

 diagram. The appearance of this indicator card is most interesting. 

 You wiU observe slight discontinuity in the rising line, and just as 

 maximum pressure is approached, the indicator begins to oscillate 

 rapidly through a small distance. These oscillations, as you see, 

 continue all down the expansion stroke, and die out gradually and 

 do not terminate until the end of the compression stroke. The 

 oscillations are about 600 to the second. The amplitude of the 

 oscillations, as you see, gradually falls, until it has practically ceased 



2 L 2 



