In the address of Jordan, President of the Societe des Ingenieurs, delivered at the annual meeting of that society in Paris, a novel explanation of the welding of iron is offered. Jordan says that welding is a phenomenon exactly similar to the regelation of water, the phenomena of rege - lation being these, that if 2 or more pieces of ice at a temperature not lower than their melting - point, or preferably at a temperature much higher than their melting - point, be pressed together, the liquid water adhering to their melting surface becomes solid at the places of contact, and thus the 2 pieces are re-frozen into one. Jordan very aptly illustrates the phenomena of regelation by the making of a snowball, telling us that this may be done when snow is at a temperature not lower than 32° F. (0° C), i.e. the freezing - point of water. Every man will remember that when the snow is very dry, and the temperature of the air below the freezing - point, the snow - flakes will not cohere without the aid of much pressure and warmth from the hand, but that with sloppy snow during a thaw, one can make a hard ice snowball with ease.

Jordan compares the making of the snowball with the welding of the iron ball, maintains that the processes are identical, and applies Sir W. Thomson's explanation of regelation to the cases of iron and platinum welding.

It appears to Prof. Mattieu Williams that the conditions of solidification in the 2 cases are not only by no means alike, but are diametrically opposite, the welding of both iron and platinum being effected at a temperature considerably below their melting - point, while the primaiy condition for the cohesion of 2 pieces of ice by regelation is that they shall be exposed to a temperature above, or at least not below, their melting - point. In order that regelation should be analogous to welding, it should take place at a temperature far below the freezing - point. Now, it is well known that under such circumstances regelation does not and cannot occur, and therefore it differs essentially and primarily from welding.

If it had been discovered that 2 or more pieces of iron, while in a furnace, raised above their melting - point and steaming into fusion, would cohere when pressed together, and that this cohesion resulted from the solidification of their liquid surfaces, in spite of the melting heat of the furnace, we should have an analogy with the regelation of melting ice, and Jordan's conclusions would be justified. Regelation means the resolidifying of a liquid, or a special cohesion in spite of liquidity; welding means a special cohesion in spite of solidity or apparent solidity. If Jordan had described them as examples of curiously opposite actions, the comparison would have been more nearly correct. We might plausibly assume that, while the pressing together of 2 pieces of wet ice produces a solidification of the surface liquid, the pressing together of 2 pieces of heated iron has the opposite effect of momentarily liquefying the surfaces of contact, and thereby soldering them together. The plausibility of this explanation is increased by the fact that pressure develops heat, and thus the welding heat might, at the surface of contact, be momentarily raised to the fusing - point, and then, on the removal of the pressure, this liquid film might solidify and thus produce the welding cohesion.

But even this theory is, in Williams's opinion, too learned. A far simpler explanation may be found, and we must never forget that when 2 or more explanations equally fit a given set of facts, the simpler is the better, and usually the true one.

In order to find a true analogy to welding, we need go no further than the vulgar "sticking together" of two pieces of cobblers' wax, pitch, putty, or clay. These are in a viscous or semi - fluid condition, and they cohere by an action similar to the transfusion or intermingling and uniting of 2 liquids. Iron and platinum pass through a viscous or pasty stage on their way from the solid to the liquid states, and the temperature at which this pasty condition occurs is the welding heat. Other metals are not weldable, because they pass too suddenly from the solid to the liquid condition. Ice, although it fuses so slowly, in consequence of the great amount of heat rendered latent in the act of fusion, passes at once from the state of a brittle crystalline solid to that of a perfect liquid. It passes through no intermediate pasty stage, and therefore is not weldable, or does not cohere like iron, etc, at a temperature below its fusing - point.

It is usual to cite only iron and platinum, or iron, platinum, and gold as weldable substances, but this is not correct. Lead should be included as a weldable metal. The two halves of a newly - cut leaden bullet may be made to reunite by pressure, even when quite cold. This is obviously due to the softness or viscosity of this metal. Outside of the metals there is a multitude of weldable substances. Glass is a typical example of these. Its weldability depends upon the viscosity it assumes at a bright - red heat, and the glass - maker largely uses this property. When he attaches the handle to a claret - jug, or joins the stem of a wine - glass to its cup, he performs a true welding process.

The chief practical difficulty in welding iron arises from the fact that at the welding heat it is liable to oxidation, and the oxide of iron is not viscous like the metallic iron. To remedy this oxidation, the workman uses sand, which combines with the oxide and.forms a fusible silicate. If he is a good workman, he does not depend upon the solidification of this film of silicate, as the adhesion thus obtained would be really a soldering with brittle glass, and such work would readily separate when subject to vibratory violence. He therefore beats or squeezes the surfaces together with sufficient force to drive out between them all the liquid silicate, and thus he secures a true annealing or actual union of pure metallic surfaces.

Cast iron or steel containing more than 2 Per cent. of carbon cannot be welded, because the compound of iron with so much carbon is much more fusible than pure iron, or than steel with less carbon, and it runs more suddenly or directly from the solid state into that of a liquid, and hence presents no workable range of weldable viscosity. (Mattieu Williams, Iron.)