The further treatment of frozen sections is given below. The relative slowness of the method of freezing, its greater cost, the necessity of the frequent replacement of the rubber bulbs and tubing, and the greater amount of trouble required by the use of ether or ethyl-chloride are disadvantages that can be avoided by the use of the carbonic acid freezing microtome. Where much work is done by the freezing method the use of the latter is advised. [Illustration: FIG. 52.—Type of freezing-microtome, for the use of ether or ethyl-chloride. Cathcart model.] [Illustration: FIG. 53.—Carbonic-acid freezing-microtome. Becker model.] [Illustration: FIG. 54.—Bardeen freezing-microtome attached to carbonic-acid gas-cylinder.] The _carbonic acid outfit_ consists of a microtome arranged so that it can be connected with a cylinder of compressed carbonic acid gas, as shown in the Aschoff-Becker or Bardeen models. (Figs 53 and 54.) The cylinders or drums containing a charge of 15-30 lbs. of the liquid carbonic-acid are furnished by the trade at reasonable rates. It is not necessary to buy the drums, as they are replaced by full ones as needed. They are provided with valves and can be fastened in an upright position or laid flat upon a table, in any way convenient for the attachment to the microtome. Connection is made between the valve of the cylinder and the object-holder of the microtome by means of strong rubber tubing which should be securely wired at both ends, or by a flexible metallic tube. The latter is preferable, as the rubber-tubing often bursts, or is so stretched by the pressure of the gas that it must be frequently replaced. Object-holders with flexible metallic tube attachment for use in an ordinary microtome are supplied by Bausch and Lomb and other firms. My personal experience makes me prefer the Bardeen freezing microtome (see Fig. 54) to all others. It is cheap and can be easily attached to the valve of the gas-cylinder, which is fastened horizontally to the table-top, as it has a screw-thread fitted to the uniform thread of the drum-valve. The object-holder can be raised or lowered between the glass-tracks on which the knife runs, so that sections of a definite thickness can be obtained. The most satisfactory knife is that of the type of a plane-bit set in a wooden handle. (See Fig. 55.) It must be well honed and stropped. The tissue, which should not be more than 5 mm. thick, is placed upon the object-holder in a drop of water, albumin fixative, or saturated solution of gum-arabic, and pressed firmly against the plate as the gas is turned on slowly and evenly. Freezing is usually accomplished in one-half to one minute. The best results are obtained by turning the gas on for about 15-20 seconds, then turning it off for several seconds; the tissue will continue to freeze; if not hard enough the gas is turned on again for 15 seconds and then turned off. If necessary this may be repeated until the tissue is properly frozen.