Nearly pure alcohol may be obtained from ordinary spirits of wine, by adding about one-third its weight of well-dried carbonate of potash, agitating the bottle, and then allowing it to stand for ten or twelve hours, during which time the potash will absorb much of the water from the spirit and fall to the bottom; the spirit may then be poured off, and fresh alkali added, and the process repeated until the potash remains quite dry, and the alcohol is then to be freed from the small portion of potash which it holds in solution by distillation in a water bath.

A far more convenient method of concentrating spirit of wine for varnish making, is that discovered by Sommering, founded upon the property of ox bladders, to allow water to pass through and evaporate out of them, but not permit alcohol to transpire, or only in a slight degree. According to Sommering, as quoted by Ure, "we should take for this purpose the bladder of an ox or calf, soak it for some time in water, then inflate it and free it from the fat and the attached vessels, which is also to be done to the other surface, by turning it inside out. After it is again inflated and dried, we must smear over the outer side twice, and the inner side four times, with a solution of isinglass, by which its texture is made closer, and the concentration of the alcohol goes on better. A bladder so prepared may serve more than a hundred times. It must be charged with the spirits to be concentrated, leaving a small space vacant; it is then to be tightly bound at the mouth, and suspended in a warm situation at a temperature of 122° Fahr., over a sand bath or in the neighbourhood of an oven. Weak spirit loses its water quicker than strong, but in from six to twelve hours the alcohol may be concentrated when a suitable heat is employed. Alcohol may also be strengthened, as Sommering has ascertained, when the vessel that contains the spirit is bound over with a bladder which does not come into contact with the liquor."

The coating of the bladder with the solution of isinglass appears, however, not to be essential to the success of the method for varnish purposes, as, upon experiment with an unprepared bladder, spirits of wine of s. p. 8.54 was brought in a few hours to s. p. 8.11, showing it to contain about 95 per cent. of pure alcohol.

Naphtha, or the spirit procured by distillation from pyro-lignous acid, and commonly known as vegetable or wood naphtha, is frequently employed instead of spirits of wine for making cheap varnishes. It dissolves the resins more readily than ordinary spirit of wine, but the varnish is less brilliant, and the smell of the naphtha is very offensive. It is therefore never employed for the best works.

The preparation of oil varnishes requires the application of considerable heat, and owing to this and the highly inflammable nature of the materials, the process is attended with considerable risk of setting the building on fire. The process should therefore always be conducted in detached buildings constructed expressly for the purpose. Owing partly to the necessity for this precaution, and the circumstance that oil varnishes are greatly improved by being kept in leaden cisterns for some months before they are used, the preparation of oil varnish is carried on almost exclusively as a separate manufacture, the details of which are greatly varied and are mostly kept secret.

In 1833, Mr. J. Wilson Neale, a varnish manufacturer of thirty years' experience, received a gold medal from the Society of Arts for a very complete description of his method of making oil and other varnishes, published in Vol. XLIX. of the Society's Transactions, from which the following directions for the preparation of oil varnishes in small quantities have been extracted: - "The copper pot employed to make the varnish, is called a gum-pot, and measures about 2 feet 9 inches in height, and 9 1/2 inches diameter externally. The bottom is hammered out of a single piece of copper, and fashioned like a hat without a brim; it is about 9 inches deep, and three-eighths of an inch in thickness. The upper part of the pot is formed as a cylinder, of sheet copper, about 2 feet 2 inches in height, and of sufficient diameter to slip about 2 inches over the upper edge of the bottom piece, to which it is firmly rivetted. A wide flange of copper, to support the pot, is also fixed just beneath the lower edge of the cylinder, and a strong iron hoop is fixed a little above the line of the rivets, to serve for the attachment of the horizontal handle, which is made as a nearly straight rod, one inch square, flattened at the end, and 2 feet 8 inches long.

"The stirrer is a copper rod about three-quarters of an inch diameter, and 3 feet 6 inches long, flattened at the one end to 1 1/2 inch in breadth for about 8 inches in length, and fitted at the opposite end with a short wooden handle.

"The ladle, which should contain about two quarts, is also of copper beaten out of the solid, and rivetted to a handle of the same metal, 3 feet 6 inches long, and fitted with a wooden handle like the stirrer.

"The copper jack, for pouring hot oil into the gum-pot, is made in the form of a pitcher, with a large handle and a wide spout; it contains two gallons. The brass or copper sieve, for straining the varnish, is about 9 inches diameter, and contains sixty meshes to the inch. The copper funnel, for straining the boiling varnish, is large enough to receive the sieve, and should be well made with lapped seams, as solder would be melted with the heat.

"The tin pouring pot, to hold three gallons, is formed exactly like a garden watering-pot, only smaller at the spout, and without any rose. This is never to be used for any purpose except pouring oil of turpentine into the varnish.

"A small broom, termed 'a swish.' used for washing out the gum-pot every time after use, is made from cuttings of cane tied to a small handle like a hearth-broom; the head is 5 inches long, and 5 inches round. This should be washed in turpentine, and kept very clean.