OF ARTS AND SCIENCES. 173 



former end is fitted, by a rubber stopper or section of tubing, a glass 

 funnel-tube provided with a stop-cock ; the end of the goose-neck 

 passes tightly through a rubber stopper in the upper end of the con- 

 densing tube. This is essentially the apparatus, but it is convenient 

 to attach to receive the distillate a small Erlenmeyer flask which 

 moves with the condenser and is joined to it, in the manner indicated 

 in the figure, by means of a thistle-tube and a rubber stopper grooved 

 to permit the free passage of air. In carrying out a distillation, the 

 liquid to be distilled is introduced into the retort either by the funnel 

 tube or previous to its -insertion, the glass cock is closed, the water 

 started through the condenser, and the retort lowered into the hot 

 paraffine, care being taken to begin the operation with the retort not 

 more than half full and so inclined that only the rear dips below the 

 surface of the bath. If the precaution to heat the retort at the start 

 in this manner be overlooked, it may sometimes happen that the 

 sudden and violent expulsion of air through the liquid will carry 

 portions of it bodily into the goose-neck, and even into the condenser. 

 With this point considered, the remainder of the operation presents no 

 difficulty, and requires little care. 



The size of the retort may be suited, of course, to the particular 

 case in hand, but for most purposes a 200 cm.^ pipette makes a retort 

 of convenient dimensions, neither too large for the distillation of small 

 charges nor too small to permit the treatment of 100 cm.^ of liquid 

 comfortably. The tube of the goose-neck should be wide enough to 

 prevent the formation of bubbles in it; 0.7 cm. is a good measure for 

 the interior diameter. It is of advantage to heat the bath to a point 

 considerably above the temperature at which the liquid which is to be 

 distilled boils, — something between 130° C. and 140° C. does very 

 well for water, and is not too high for methyl alcohol, — and under 

 such circumstances, and when the retort is entirely submerged, it 

 often happens that evaporation takes place with extreme rapidity from 

 the surface of the liquid in perfect quiet without actual boiling. 



With such an apparatus the following experiments were made. 

 The boric acid was weighed, as before, in solution, and to bring the 

 condition of the experiment to that of an actual analysis, 1 grm. of 

 pure sodium hydrate was added in solution, nitric acid or acetic acid to 

 acidity and a little more, and the whole was introduced into the retort 

 and distilled to dryness. 



In those experiments in which nitric acid was employed, the methyl 

 alcohol was introduced upon the residue thus dried in six successive 

 portions of 10 cm.^ each, and distilled to dryness ; but in order to break 



