TWENTY-THIBD ANNUAL MEETING. 



165 



A NEW GLASS SIPHONING EXTRACTION APPARATUS. 



J. T. WILLARD AND G. H. FAILYEB. 



At the meeting of the Academy in 1885 we had the pleasure of describing what 

 still seems to us the simplest percolating extraction apparatus constructed wholly 

 of glass.* We now describe a siphoniiuj apjjaratus which it would seem can hardly 

 be simpler, and which after nearly a year's use we can pronounce entirely satisfactory 

 in its operation. 



It may not be out of place to say a few words in justification of the invention of 

 another extractor, notwithstanding the many forms already proposed. In the first 

 place, we are led to reject any extractor which makes cork connections. The ex- 

 treme difficulty of completely freeing cork from its resinous constituents can only 

 be appreciated by one who has tried it faithfully for a hundred hours or more, plac- 

 ing the properly perforated corks in a large Soxblet extractor and changing the 

 flask occasionally and constantly finding an appreciable extract. The writers have 

 also observed an old, well-used cork suddenly exude a considerable quantity of resin 

 on the upper end of the cork during an extraction. The more perfectly a cork is 

 freed from its resin the less fitted, mechanically, does it become, losing its elasticity 

 and becoming brittle. In the second place, a six)honing apparatus must be prefer- 

 able to a percolating, because we are certain that every particle of the substance is 

 soaked by fresh solvent every time the container is filled, while in a percolating ap- 

 paratus we can never be quite sure that the solvent is not chiefly flowing down chan- 

 nels of least resistance. 



The Soxblet extractor is sometimes made wholly of glass, and in that form will 

 doubtless do good work. Ours, however, has the advantage of greater simplicity, 

 and consequently costs less, as well as being less liable to fracture. By 

 reference to the cut it wiU be seen that the extractor consists of four 

 parts, viz.: the flask, the containing tube, the condensing tube, and the 

 siphon. The flask is small, holding only about 100 cc, and weighing 

 about 20 grams. The containing tube is ground into the neck of the 

 flask, and has a pocket on one side, in which the substance, wrapped 

 in filter paper, is placed for extraction, and into which the short arm 

 of the siphon dips. The condenser tube expands into a stopper for 

 the containing tube, and at the lower end is drawn out and ground off 

 in such a way as to direct the condensed solvent into the pocket con- 

 taining the substance for analysis. The upper part of the condenser 

 tube passes through a cold-water jacket of some kind; in ours a tank 

 large enough for half a dozen extractors. The siphon is made of smaU 

 tubing. The internal diameter must not be over one millimeter, and 

 at the bend it is better to have it even less. The capillary attraction of 

 the inner wall of the tube raises the liquid sufficiently to bring the si- 

 phon into action before the pocket is fuU enough to run over. The 

 upper end of the short limb must be bent sharply away from the side 

 of the pocket so as to cut off the capillary action between the outside 

 of the siphon and the side of the pocket against which it touches. If 

 this is not done, the liquid will be drawn over continuously without 

 emptying the pocket, and all the advantages of a siphoning ap- 

 paratus will be lost. If, however, these details be met in the con- 

 struction of the siphon, it will act perfectly, and the apparatus 

 will need no attention after the flame under the water bath is adjusted. The ap- 



* Transactions Kansas Academy of Science, Vol. X, p. 20 ; Am. Chem. Journal, Vol. 8, p. 73. 



