EXPERIMENTS IN REFINING WOOD TURPENTINE. 29 



but the vapors ascend from one chamber to another through a pipe 

 the center of the plate, are deflected downward by a hood over 

 the pipe, and forced to boil their way through the liquid on each 



>late by the steam which enters the still through the pipe 3. The 

 arrows indicate the course of the vapor in the still. The vapors 



>oil from chamber to chamber, becoming purer as they ascend until 



hey reach the top chamber of the column, from which they are 

 delivered through the pipe 4 into the cooler condenser 5, where they 

 are reduced to a liquid. This cooling and condensing is effected by 



jirculating cold water through the tubes around which the vapor 

 passes. 

 The condensed vapor, turpentine, and water is drawn off at the 



>ottom of the condenser and allowed to flow through the test box 6, 

 where it can be examined. The steam pressure within the apparatus 

 registered by a pressure gauge not shown. The turpentine and 

 water after passing through the test box is run into the separator 7, 



rom which the turpentine may be run to storage tanks. 

 The general plan of the experiment was to draw off samples from 



ime to time during the distillation, noting the temperature of the 

 vapors escaping to the condenser and the relative volumes of tur- 



)entine and water in the condensed distillate, and making an exam- 

 ination of the samples, determining specific gravity, refractive 

 index, initial distilling temperature, and behavior on distillation 

 when 150 cc were distilled from an ordinary distilling flask, using an 

 emergent stem thermometer. All of the turpentine obtained during 

 a distillation was not run into the same tank, but at various prede- 

 termined periods it was directed into other tanks, and at the close of 

 a distillation each fraction was measured, and the relative volume of 

 water and turpentine determined. The residue of heavy oil was also 

 measured, and samples from each tank were examined as just out- 

 lined. The results of the distillation and data obtained on the dis- 

 tilling oils from time to time during the process are given in Tables 

 4 and 5. 



These data show the effectiveness of the separations at intervals 

 during distillation and are in all respects similar to data obtained in 

 the distillation of other closely related oils with steam in a column 

 still. The first portions distilling consist almost entirely of light oil, 

 but as distillation proceeds the proportion of light oil in the distilling 

 oils decreases while the proportion of heavy oil increases until distil- 

 lation is discontinued. The fact that the proportion of heavy oil 

 begins to increase when the distillation is about half over is proof 

 that large quantities of heavy oil were in the crude product. The 

 first or A fractions of all distillations in the beer still were mixed 

 together, and the A fractions of all distillations from the refining 

 still were mixed together; each mixture was then again passed 



