FLAME TEMPERATURES. 169 



gas, 8| parts of air. It will be noticed that the proportion of 

 oxygen is sensibly less than that required for perfect com- 

 bustion. 



The luminosity depends on the compression of the gases 

 or the air. Hydrogen burning in oxygen at ordinary pressure 

 gives a flame hardly visible at all ; with a pressure of 20 atmos- 

 pheres it becomes quite luminous. Arsenic in burning pro- 

 duces quite a luminous flame at ordinary air pressure; but 

 hardly any in rarefied air. The same is true of carbonic 

 oxide and other gases. The luminosity seems to be in direct 

 proportion to the pressure. 



Luminosity seems to be greater with those substances 

 which on burning produce dense vapors. Hydrogen and 

 chlorine produce a vapor twice as heavy as water and the 

 luminosity is much stronger than with the oxygen-hydrogen 

 flame. Carbon and sulphur also produce heavy vapors and 

 much light. Phosphorus burning in oxygen produces the 

 dense heavy phosphoric anrrydride and this is accompanied 

 with an almost blinding light. 



The length of the flame ordinarily depends on the quantity 

 of hydrogen, and consequently the hydrocarbons contained 

 in, or generated from, the body consumed. With fuels con- 

 taining high hydrocarbon percentages, flame of almost any 

 desired length can be produced. This is especially the case 

 with gases. 



The theoretical temperature of combustion, and hence of 

 the flame, may be calculated by dividing the heat units pro- 

 duced by the specific heats of the products formed. Of course, 

 these theoretical temperatures are never reached in practice, 

 but they serve as aids in determining the value of fuels for 

 certain purposes. 



A few typical examples of these calculations will be given. 



I. Hydrogen. Hydrogen burnt in oxygen produces 

 29000 heat units (water considered as vapor); the specific 

 heat of the aqueous vapor produced is 0.475. The hydrogen 



