Measurement of Gas-Engine Temperatures. 95 



10 feet per second was flowing. The wires were at right 

 angles to the length of the box and were heated by an electric 

 current, the resistance being measured at the same time. It 

 was found that the 4 mil wire lost heat about 50 per cent. 

 faster than the 2 mil wire at the same temperature. In the 

 case of the larger wire, the rate of loss was '0028 watt per 

 centimetre length for every degree centigrade by which the 

 wire was hotter than the air current at a distance from it, 

 and this relation held good up to a temperature of 600° or so, 

 when the radiation became important. In the suction stroke 

 of the gas-engine the loss is considerably greater, being about 



= -0038 watt per centimetre per degree, a result easily 



explained by the more violent motion of the gas. 



How great is the effect of convection is readily seen by 

 comparing the figures just given for the rate of loss with that 

 obtained by calculation from the thermal conductivity of air 

 on the assumption that there is no convection. If a cylindical 

 wire of radius r, and at temperature 0, be surrounded by a 

 concentric cylinder of radius r at temperature # , the inter- 

 vening space being filled with air at rest, the rate of flow of 

 heat from the wire per centimetre length will be 



2ir£(0 1 -0„)- 



log e a 



1 I 



where h is the thermal conductivity of air. Taking k as 

 •00005, r ± as ~ inch, r as 2 inches, we get for the rate of 

 flow 4*15 x 10 -5 calories per second, or 1*74 xlO -4 watts. 

 This is only l/16th of the rate found in a wire of the same 

 size placed in a current of air flowing about 10 feet per 

 second, the dimensions of the surrounding enclosure being of 

 the same order as a cylinder of 2 inches radius. 



I have great pleasure in acknowledging the help that I 

 have received in this investigation from Messrs. A. L. Bird 

 and A. R. Welsh, two students at the Engineering Laboratory, 

 Cambridge. They made and reduced the whole of the 

 observations on which the work is based. 



Engineering Laboratory, Cambridge, 

 October 20, 1906. 



