90 Prof. B. Hopkinson on the 



5\3 ohms, and the current is 0*765 amp. Heat is therefore 

 supplied to the wire by the current at the rate of 3*1 watts. 

 The rate at which the wire is losing heat to the gas along 

 PP' is accordingly 3-1 + 2-16 = 5-26 watis. This is\(0"-0). 

 Along Q Q' the rate of loss of heat is 1*61 watts, and this is 

 the value of \(6 r —6). The difference, viz. 3'()5 waits, is 

 equal to X(0"-0')« The mean value of 6" -6' over the 

 range is 40°' 5, whence 



X=Mj>=-09. 



40-5 



This is the rate (in watts) at which the wire loses heat to the 

 gas per degree difference of temperature. The difference 

 of temperature between wire and gas along Q Q' is therefore 



- m =17°*9 C, and the gas temperature at the middle point 



of the range (615° crank-angle) is 12° C. Thus the point li 

 on the gas curve is found. 



The wire temperatures are probably correct within 1°. If 

 all four temperatures were in error to that extent, and if the 

 errors were disposed in the most unfavourable manner, X 

 would be wrong by "0085, or about 10 per cent, of its value, 

 and the error in the correction 6' — would be about 3°. It 

 is unlikely that all the temperature errors will combine to 

 produce the maximum error in the result, and probable that 

 this particular gas temperature is correct to within one or two 

 degrees. At other points the gas temperatures are not so 

 accurate ; the possible error increases with the slope of 

 curve B, and where that curve departs greatly from a straight 

 line a smaller time interval must be taken to get the slope. 

 Thus at the end of the compression or the middle of the 

 expansion strokes, when the correction amounts to nearly 

 100°, the error may amount to 10 or 15 degrees. 



The pressure in the cylinder throughout the cycle is shown 

 by the dotted curve in fig. 2. The pressure was obtained 

 from indicator diagrams and (during the suction and exhaust 

 strokes) by a water-pressure gauge which could be momentarily 

 opened to the cylinder at any point in the cycle. 



Several points of interest appear in a study of the curve of 

 gas temperatures, and a comparison of it with the pressure 

 curve. We may discuss these in order, beginning with the 

 suction stroke [510 °-720° crank angle]. During the suction 

 stroke the pressure in the cylinder falls considerably below 

 atmospheric; but the work done on the air in forcing it through 

 the constricted opening of the inlet-valve is almost wholly 



