PROCEEDINGS OF SECTION B. 



105 



The gas was found to be faintly luminous. Sample "A" was 

 obtained by placing the retort in a furnace, the temperature of which 

 had been previously raised to about 700° C. In the other two cases 

 the furnace was preheated to about 200° C, and then raised gradually 

 to the maximum temperature, the rate of increase being roughly twice 

 as rapid in the second case as in the third. 



The volumes of the gas collected were not corrected to normal 

 temperature and pressure, but fairly correspond to those which might 

 be expected under working conditions, as the main bulk of the admixed 

 ^ir was introduced after the measurement of the volume. Under the 

 conditions of experiment it would have been difficult to obtain air- 

 free samples, and as the errors introduced are in the right direction, 

 that is lead to a conservative estimate of the calorific value of the gas, 

 no attempt to obtain purer samples was made. 



Table I. 



Acid. — The first liquid distillate was a dilute aq\ieous solution of 

 acetic acid contaminated with tarry substances. In order to discover 

 whether any other acid substances were present, this solution was 

 evaporated to dryness after neutralization with sodium hydroxide. 

 No acid other than acetic was found, except a small trace of phenol. 



The usual procedure was as follows : — The quantity of acid present 

 was determined by titration with standard sodium hydroxide, phenol 

 phthalein being used as indicator. In some cases, however, the 

 solution was neutralized with calcium carbonate and fractionally dis- 

 tilled with a rod and disc dephlegmator. It was found that a very 

 small proportion of the liquid distilled over between 96° and 100° C. 

 This was taken to show that if methyl alcohol be present the amount 

 is quite negligible. 



Basic Substances. — After titration as above, the solution was made 

 alkaline with strong sodium hydroxide solution, and the distillate 

 collected in a known volume of standard acid, the excess of which was 

 estimated by back titration with standard alkali. This solution was 



