X. PRESIDENTIAL ADDRESS. 



In the internal combustion engine the working fluia 

 is air, which is heated by the direct combustion of the 

 fuel inside the cylinder to as high as 1700 degrees or 1800 

 degrees Centigrade. Since the working parts could not, 

 for obvious reasons, be allowed to attain this tempera- 

 ture, the positive cooling of tlie cylinder walls is a prac- 

 tical necessity. In the steam engine the temperatures 

 are easily dealt with, and the abstraction of heat from 

 the cylinder walls is to be avoided as being merely a 

 source of waste. 



Since increase of cylinder dimensions means a de- 

 crease in the ratio of area of cylinder walls to cylinder 

 volume, it will be seen that, whereas this condition is 

 desirable in a steam engine, in a gas engine it merely 

 leads to practical difficulties in keeping the temperature 

 of the working parts within a practical limit. For in- 

 stance, when the diameter of a gas engine cylinder is 

 greater than about 20in. it l)ecomes necessary to water 

 cool the moving piston, whereas in smaller sizes water 

 jacketting of the cylinder Avails is quite sufficient. In 

 large cylinders also the increased thickness of metal re- 

 quired to resist the explosion pressure leads to liability 

 to severe internal sti'ains, due to unequal expansion of 

 the metal, and pistons and cylinders have frequently 

 cracked from this cause. In addition to the above 

 trouldes with large gas engines, the fuel used in the 

 majority of the larger size units was either coke oven 

 gas or blast furnace gas, and it was not properly realised 

 that efficient purification of such gas was necessary to 

 avoid fouling of the engine. 



The result has been that the large gas engine has in 

 the past gained an unenviable reputation for unrelia- 

 bility of operation. Profiting, however, from past experi- 

 ences, designers have now been able to produce large 

 gas engines for various classes of land service which ap- 

 proach very closely the reliability of operation of good 

 steam engines. It must not be supposed, how^ever, that 

 the one class of engine will satisfy all classes of service. 

 Many of the failures to secure satisfactory service have 

 been due to the attempt to apply the one class of engine to 

 all classes of work. Now, in steam engine practice, very 



