drawn off in reducing the pressure to atmosphere. 

 The curve for Expt. 7 shows the retardation when 

 sodium carbonate was used. The extraction of 

 rosin was in this case incomplete. The vertical 

 portion of this curve indicates the turpentine 

 fraction recovered during final cooking for pulp. 



Another outstanding feature is the remarkable 

 similarity of the curves, in spite of the differences 

 in treating the wood. In the three experiments 

 where caustic soda was used it will be seen that 

 the curves almost converge at the two hour mark. 

 This indicates that it makes little difference wheth- 

 er distillation is commenced as soon as pressure is 

 is evidenced or whether relieving is postponed a little. 



The gradual flattening of all the curves and the 

 marked retardation in turpentine evolution in 

 Expt. 7, where solution of the rosin was slow and 

 incomplete, indicate pretty clearly that disintegra- 

 tion of the rosin is necessary for efficient turpentine 

 recovery. This is the difficulty experienced in 

 the ordinary steam distillation process, where the 

 fine resinous material must be subjected to the 

 action of steam for 3-10 hours and even then with- 

 out the recovery of the last traces of oil. In a degree, 

 then, the curves represent the speed of alkali 

 penetration of the wood. Evidently, an important 

 phase of the distillation is to ensure rapid removal 

 of the turpentine from the digester during the 

 last stages of the steaming period, so that the 

 operation can be discontinued as soon as the rosin 

 is dissolved and the last of the oil thereby liberated. 



The amounts of steam necessary to carry over 

 the various turpentine fractions are important 

 from a practical view-point. In nearly all cases 

 the first 50 per cent of the oil was distilled 

 with less than its own volume of condensed water- 

 vapor. The next 25 per cent required from 1 to 4 

 times its volume of aqueous distillate, the ratios 

 in the most representative experiments (4, 5 and6) 

 being 1.2, 1.3, and 1.0 respectively. The fraction 

 from 75 to 90 per cent required ratios of from 7 to 10 

 under properly regulated conditions. The figures 

 for the final 10 per cent are not significant, be- 

 cause of the large amount of steam removed at 



41 



