11 



lars instead of through the tube itself. The bulbs at the upper 

 ends of the U tubes prevent them from falling, and the bends 

 arc made in order to hold a small amount of liquid, and thus act 

 as a trap to prevent the escape of vapors through the tubes in- 

 stead of through the gauze. Each trap was one-fourth inch shorter 

 than the one below it, since it was found advisable to have succes- 

 sively heavier water seals from top to bottom. Each constriction 

 9 as made smaller than the one above it, with the exception of 

 the two at the top, in order to allow the collars to be easily placed 

 in position when the dephlegmator is being constructed. A strip of 

 40-mesh brass gauze a three-eighths of an inch wide was carefully 

 wrapped about each tube until the size of the collar was sufficient to 

 fit its particular constriction snugly. The U was then dropped care- 

 fully into the large tube and gently forced into position with a glass 

 rod. A snug but not tight fit is desired to prevent the tubes from 

 blowing out of position while the apparatus is in operation and at 

 the same time to permit of easy withdrawal for cleaning. To facili- 

 tate removal, a loop of copper wire is fastened to the top of the tube 

 into which a hook can be inserted. In this way each section can be 

 readily removed. It is a matter of some skill to set up the dephleg- 

 mator, but when the apparatus is once in position it is ready for a long 

 series of determinations. Cleaning, in the case of turpentine distil- 

 lations, can be done by ordinary solvents, such as alcohol and ether, 

 without taking the apparatus apart. 



Before use each time it is necessary to see that the bent tubes 

 inside the dephlegmator are filled with water. This is done by pouring 

 water into the top and allowing it slowly to run out. 



OPERATION. 



The operation of the apparatus is simple. The boiler is about 

 half filled with water, and the separatory funnel G is entirely filled. 

 The flask A, previously dried and weighed, is placed on a balance 

 capable of weighing accurately to 10 mg., and exactly 500 grams of the 

 sample of turpentine is poured in. The flask is then replaced in the 

 position shown in the diagram. A sample of the original oil should be 

 reserved for comparison with the several fractions of the distillate. 

 The burner under the boiler is then lighted, and when the steam is 

 freely escaping through the valve J, the valve is closed to allow the 

 vapor to pass into A and the burner turned down to a point deter- 

 mined by experience. It is necessary to have the distillation carried 

 on at a slow and uniform rate. Two drops of distillate per second 

 is the rate that has been found to give the best results. In order to 

 keep this rate constant all heat conditions must be carefully regulated. 

 The flame under A should be small and low to prevent excessive con- 



Platinum could, of course, be used to distill liquids that would attack brass. 

 [Cir. ir.ii] 



