If a white sediment forms in the solution, the sample should be at once rejected.

In the application of a preservative coating to iron, Prof. Lewes directs, first, thorough scraping and scrubbing from all non-adherent old paint and rust. New iron should be pickled with dilute acid to get rid of every trace of mill scale; the acid to be neutralised afterward by a slightly alkaline wash, and this again to be washed off by clean water. Under these conditions, and given a composition of good adhering properties, but little apprehension need be felt with regard to the ravages of corrosion, the chief remaining risks being from abrasion or other mechanical injury to the composition, coupled with. improper contituents in itself, (Scient, (23) For several years, G. W. Gesner, of New York, has been eiperimenting with a process for giving articles of iron and steel a rust-proof coating. Plant has been established at South Brooklyn, which has been in operation for some time past. The accompanying illustrations, Figs. 266, 267, and 268, show its construction. It consists substsntially of a bench of two ordinary gas retorts placed side by aide in a furnace heated by a grate. Each retort is heated to a temperature of 1000-1200° F., as may be determined by the character of the articles to be treated.

The latter are introduced by means of a crane and pulley, care being taken that they do not touch one another. After closing and testing the retort, the heating continues for about 20 minutes. Then steam is introduced into what Gesner calls a " hydrogen generator," shown in Figs. 267 and 288. It is a simple pipe, open at the rear end. Gesner claims that in the passage of the steam through this generator hydrogen is generated, which fills the retort. This operation goes on for 35 minutes, at the end of which time.

Gemer's rust-proofing furnace.

1/2 pint of naphtha is permitted to flow into the retort for 10 minutes. The flow of hydrocarbon is then stopped, and the steam which has been allowed to enter the generator daring the whole operation is continued for 15 minutes longer. The whole time employed in the operation is therefore 1 hour and 20 minutes. The "purging-pipe," which dips into an open vessel of water, as shown in Fig. 266, to the depth of 1 1/2 in. carries off any excess of gases produced in the operation. In cases where articles treated are ornamental, such as art hardware, they are given a bath of cold whale oil or paraffin oil to render them more even in tone. In other articles no oil is used. The plant now established at South Brooklyn is rated at a capacity of 6 tons per day of boiler tubes 7 1/2 ft. long, or 2 tons of ornamental hardware, the rate of production of treated goods depending upon the time required for handling them. The average cost in America of fuel per day is reported by Gesner to be 7s., including coal for the boiler.

Gunner's rust-proofing furnace.

To substantiate his claim that hydrogen has a function in the creation of a rust-proof coating, Gesner quotes the following analysis, made by Stillman & Gladding, of New York, of a sample of the surface of cast iron prepared by the process: Carbon, 1.01 per cent.; hydrogen, 0.22 per cent.; sand, 6.70 per cent.; and iron, 66.10 per cent. The chemists add that the iron is present as metallic iron and as oxides of various constitution.

In order to determine whether the treatment had any adverse effect upon the strength and resistance of wrought iron and steel suitable for boiler, ship, and bridge purposes, a series of tests were made by B. H. Coffin, in charge of the testing department of Henry Warden, Germantown Junction, Phila-'delphia, Pa. We quote as follows from this report: -

"5 test pieces of iron were cut from a single plate 1/2 in. thick, and 5 more similarly from a 3/8 in. steel plate. These were machined to suitable sizes for the standard 8 in. test piece, giving a section of about 9.71 sq. in. for the iron and 0.51 sq. in. for the steel. Three of each of these sets were forwarded to Gesner for treatment, who retained one and returned the remainder. The tests were made with a 200,000 lb. Olsen machine, and the measurements with Brown & Sharp's micrometer gauges, and are believed to be accurate.

The pieces were gauged both before and after treatment, and showed no change. The tests show practically no effect whatever upon the iron, with the exception of a slight decrease in the elongation. As the reduction of area is not fully ascertained, it is impossible, without further evidence, to say whether or no the ductility is affected. At any rate the ductility being so low, this small reduction, if proved to exist, would be of comparative unimportance in affecting the value of the metal. The steel is benefited. The annealing undergone during the treatment has softened it to some extent; it has lost about 5 per cent, in strength but gained 5 per cent, in elongation. This metal originally would not have come up to specifications, being insufficient in stretch. The treatment has not reduced the tensile strength below the assigned limit; at the same time it has brought the elongation up to requirements. Pieces of both iron and steel were bent cold to an angle of 45° without showing any fracture or scaling of the treated surface.

The following are the results of the tests: -

A large variety of iron and steel goods have- been already treated in quantities by the South Brooklyn Rust Proof Iron and Steel Works, among them being builders' and art hardware, roofing shingles, stove fittings, pipe and pipe fittings, parts of water meters, steam radiators, pistols, and other articles. The colour produced is a dark blue. (Iron Age.)