In a case like this there is nothing to be got by turning and twisting the thing round so as to expose alternately its various parts to the greater heat. In imagination one can see what occurs; by a process of distillation and condensation the same kind of transference of moisture from the hotter to the colder parts takes place as is seen to occur, for instance, when the attempt is made to dry a damp flask under similar circumstances.

But even this picture by no means exhibits the whole of the perplexing difficulties of tracing and controlling the air currents in an ordinary air-bath. They are so erratic, so fortuitous, so delusive, and subject to such uncontrollable shiftings, that, practically speaking, it is a question whether thorough drying in such a case is possible without destruction of the substance. At all events the chances in favour of it are so slender as not to be relied upon.

Reflecting upon these difficulties, the theory of a perfect method for air-drying suggested two things as indispensable.

1. A constant current of pure dry air brought to the desired degree of temperature before admission into the drying chamber.

2. A regulated source of heat.

Both these conditions to be under perfect control.

I place current first as being the more important item, and the one to which hitherto very little attention appears to have been paid so far, at all events, as regards the temperature of the current when admitted into the drying chamber. As for the source of heat, this has been brought fairly under control by previous experimenters by the use of one or other of the several thermostats, though 1 was unable to discover one quite suited to the purpose.

To meet these requirements I have designed the instrument shown in Fig. 18. It consists essentially of a double jacketed cylinder with air passages, so contrived as to compel the ajr used for the drying to circulate between the jackets before its final admission into the drying chamber, in such a manner as to ensure that its temperature shall be raised to the required degree before it is allowed to come in contact with the thing to be dried. Moreover, as a secondary, but still very important consideration, the inlet of the air-supply is so placed that the air used for drying is kept as distinct as possible from all contact or admixture with the products of combustion, such products naturally being loaded with watery vapour and C02. The pure air thus heated before entering the drying chamber,is evenly diffused over the whole area of the bottom of the bath, between it and the perforated false bottom which forms the floor of the drying chamber; the air then ascends bodily as a solid cylinder, and escapes by the tall chimney of the domed glass cover. In this way vigorous circulation of dry hot air is constantly main-tained, which effects a, rapid and uniform drying, such as no ordinary air-bath can accomplish. We will now turn to a consideration of the source of heat and its regulation.

This consists of a Bunsen air-burner placed in the cavity under the bottom, but the heat is first received upon a solid disc of metal, separated by a sufficient space from the bottom to prevent the heat of the flame being transmitted direct to the bottom of the bath, the object being to avoid any localisation of the heat. Moreover, the mass of the metal disc, besides acting as a distributor, also serves as a reservoir of heat and assists in maintaining the equality of temperature, but this equality is chiefly provided for by an entirely new form of thermostat. It is to be observed that within the two jackets already spoken of is placed an annular copper vessel, which forms the boundary wall of the drying chamber. This is a cylinder composed of two thicknesses of thin sheet copper, enclosing an air-space of 5 mm. wide, 98 cm. circuit, and 22.75 cm. in height. It is securely closed top and bottom, and has a capacity of 1100 cc. This constitutes the heat regulating chamber or thermostat, the cavity of which is connected up to a U tube, having mercury in the bend.

Gas is admitted on the other side of the U, and by means of an arrangement, such as is usual in a gas thermostat, depression of the mercury in one limb cuts off the main gas supply, which can then reach the burner by a small bye pass only. By means of a screw at the top of the other limb of the U, air can be admitted into or allowed to escape from the regulator. With the rise or fall of the temperature, and the consequent expansion or contraction of the air contained in the regulator, pressure is exerted or withdrawn from the surface of the mercury, which is thereby forced down the one limb and up the other.

Adam's air bath.

Adam's air bath.

The reference letters indicate as follows:-A, Diaphragm completely separating the drying from the combustion chamber. B, Perforated false bottom. C, Outer jacket. D, Inner jacket. , Copper regulating chamber or thermostat. F, Baffle plate. G, Apertures in jacket C giving admission to air for combustion. H, Apertures for the passage of air between the jackets for drying. I, One of three apertures for escape of products of combustion. K, Apertures in inner jacket D for passage of drying air. L, Burner. M, Thick metal plate for receiving heat of the flame. N, Mercury U tube. O, And its connection with copper regulator. P, Screw whereby the degree of heat is regulated. R, Gas supply. S, Tube with bye-pass,

I have already mentioned the capacity of this copper regulator is 1100 cc, the coefficient of expansion for one degree Centigrade being '00367, the alteration of volume for a single degree of temperature at boiling point will be about 3 cubic centimetres (2.95). It is therefore plain we have here a means of regulation of the temperature of extraordinary sensitiveness, and accordingly we find we can command what practically amounts to a fixed temperature at any desired degree, and seeing that the copper regulator entirely surrounds the drying chamber and that the whole of the air employed in the drying process must of necessity sweep both its surfaces, exterior and interior, amounting to nearly half a (•4459) square metre in extent, it follows that no local currents can interfere with the accuracy of its workings. You will agree, I am sure, that this a grand point.