194 Prof. W. M. Thornton on the Limits of 



line with the others if its combustion in the explosion wave, 

 even in dilute mixtures, is to CO and not to C0 2 . In this 

 case n is 2 and Ln L =38. It has been shown previously 

 that the upper limit is inversely proportional to the number 

 n of oxygen atoms burnt. At the lower limit oxygen is in 

 excess, but it is now found that the product Ln L of the per- 

 centage of gas and of the number of oxygen atoms in the 

 lower limit mixtures, given in the last column of Table II., 

 is also constant. Since both TJn and Lw L are fixed their 

 ratio is constant. But in all homologous series of gases n/n^ 

 is sensibly constant. It follows that the ratio of the upper 

 to the lower limits of inflammability should be nearly constant. 

 In the paraffins this ratio increases slightly, in the ethylene 

 series it decreases, as shown in Tables III. and IV. 



4. Since the product of the percentage L of combustible 

 gas in the lower limit mixture and of the oxygen atoms 

 present per molecule of combustible gas is constant, the 

 number of oxygen atoms is the same in unit volume of the 

 lower limit mixtures of all inflammable gases. These atoms 

 are not all burnt, but as they are all raised to the temperature 

 of ignition it follows that they are all activated, and that 

 the intensity of pioneering ionization or activation of oxygen in 

 the advancing wave-front is the same in all gases at the point 

 where self-ignition begins. 



The percentage U is a measure of the number of com- 

 bustible molecules in unit volume of the upper limit mixture ; 

 and since the heat of combustion of each molecule of a 

 hydrocarbon can be shown to be approximately proportional 

 to r/, the product JJ?i is proportional to the total heat of com- 

 bustion of unit volume, and this is the same in the upper limit 

 mixtures of all groups of gases in which TJn has the same 

 value. 



Both the upper and lower limits of inflammability can 

 then be considered as controlled by the heat liberated in the 

 reaction, and this would bring the facts stated into line with 

 current views. It is not, however, the heat set free that 

 controls the oxygen that can be present and so decides the 

 percentage of gas, but the oxygen that controls the heat, so 

 that behind the thermal effects there are the conditions of 

 Tables I. and II. Inflammation can only occur when certain 

 numerical relations exist between the oxygen and gas 

 molecules. 



5. It is possible from the above to predict the limits of 

 inflammability of certain groups of compounds t with fair 

 hope of accuracy. The heavier paraffins as far as octane 



