to prevent surging over of the liquor. After four 

 hours heating the distillate measured 165 cc., 

 containing 28 cc. of water white, sweet-smelling 

 turpentine, which represented a yield of 27 gallons 

 per cord. The recovery of volatile oils was in- 

 complete and undoubtedly the distillate contained 

 only the lighter fractions. At this stage the chips 

 seemed pretty well extracted. A sample of the 

 alkali extract gave a very clean, white precipitate 

 of rosin soap on cooling, and the supernatant 

 liquor was transparent and light cherry red in 

 color. 



The main portion was heated for some time lon- 

 ger under about Sin. vacuum, and distillation pro- 

 ceeded much more rapidly at 85 90. A small 

 additional amount of turpentine was obtained 

 bringing the yield up to 30 gallons per cord. The 

 liquor assumed a decidedly darker color, but the 

 rosin soap precipitate was much cleaner than in 

 the previous experiments. 



EXPERIMENTS WITH SODIUM CARBONATE 



Sodium carbonate was tried as an extracting 

 agent in the hope that the attack of the lignocellulose 

 would be minimized and solution of the rosin ac- 

 complished satisfactorily. 



1. Atmospheric Pressure 



Shavings were readily extracted by dilute solu- 

 tions of sodium carbonate at temperatures ap- 

 proaching the boiling point. The action was slow at 

 lower temperatures, but when shavings were added 

 to a boiling solution of carbonate there was in- 

 stantaneous action, evidenced by vigorous frothing 

 due to evolution of C02 and turpentine. After 

 boiling for three hours the solution was cooled, giv- 

 ing a clean-looking rosin soap precipitate and a trans- 

 parent reddish liquor. It was found that the ex- 

 tracts were decidedly freer from humus than in 

 the case of NaOH, but all the solutions were 

 brown and the problem of humus contamination 

 was only one of degree. 



2. Heating Under Pressure 



To determine the effect of pressure, 3000 grams of 



23 



