In case it is desired to weigh the extract it may be done directly by weighing it in the test-tube T after drying in the usual way at the end of the extraction; or a glass flask H, made to fit freely into the test-tube, may be used, in which case a little mercury is poured into the bottom of the tube to seal the space between H and T. To prevent spirting of the substance in H, or projecting any of the extracted material without or against the bottom of the crucible G, the funnel represented by the dotted lines in the right hand section may be used. Heat may be applied to the test-tube either by hot water, or steam, or by a bunsen, which permits of the flame being turned down to minimum proportions without danger of burning back. When the test-tube alone is used it is advisable to first put into it some fragments of pumice stone, particles of platinum foil, or a spoonful of shot, to prevent bumping of the liquid when the lamp is used as the source of heat. Any air which the apparatus contains is pushed out through F when the boiling begins, the tube F not being closed until the vapor of the liquid has reached its maximum height. With cold water in the condenser the vapor of ether very rarely reaches above the lower compartment and the vapor of alcohol rarely above the second. When the plate P is accurately turned so as to fit the ground surface of the mouth of T, it is found that ten cubic centimeters of anhydrous ether or alcohol are sufficient to make a complete extraction, and there is not much loss of solvent in six hours. The thickness of the asbestos film in G, or its fineness, is so adjusted as to prevent too rapid filtration so that the solvent may just cover the material to be extracted, or, after the material is placed in a crucible, a plug of extracted glass wool may be placed above it for the purpose of distributing the solvent evenly over the surface of the material to be extracted and of preventing the escape of fine particles. [Illustration: FIGURE 24. IMPROVED COMPACT EXTRACTION APPARATUS.]