DISCUSSION.

Mr. Wm. Kent.-Please explain how a ring or hoop can be electrically welded from a bar or strip of metal. What prevents the current going around the ring through the solid metal, rather than across the joined ends?

Mr. Woodbury.-The resistance of a conductor is inversely as the length, other things being equal. In welding a ring by this process, a bar is bent to the form of a circle or other closed curve, the clamps holding the bar near to its juncture and the distance between the clamps will be as small as practicable; for example, it may be of the circumference of the circle. Then disregarding the varying conditions of conductivity, at the weld during the process of welding, 8 of the electricity will pass from one side to the other through the point of juncture, and only of it will go around the longer side. As I stated in the course of the paper, the heating effect is proportional to the square of the current and other conditions, and therefore the union between the joined ends of the rod will be heated to a welding point, while the rest of the loop will not be appreciably warmed. I have in my hand a small ring of an inch in diameter, welded by electricity, and a number § of them of various sizes are among the examples of welding shown. It is in practical use for much larger work of this nature, such as spinning rings and carriage tires being joined by this process.

Mr. Oberlin Smith.-I visited the laboratory of the Thomson Welding Company a few weeks ago, and staying there and working over their machines for an hour or two, I saw several very interesting experiments tried. Among others which I tried myself was the welding or soldering together of two pieces of ordinary tin plate, each piece perhaps the size of half a dollar. It is really welding, of course, but the metal has to be brought to a heat due only to the melting temperature of tin, rather than iron, because it is the tin surfaces that are melted together rather than the iron beneath. Thus we may call it either soldering or welding, although there was no artificial solder used-nothing but the ordinary coating of the tin plate. It seemed to unite perfectly and make an excellent flat joint. The whole surface was pressed together, or sweated together, as the term is with tin men. The field in which I was working was the soldering of certain articles together, if possible, by electricity, which now are soldered by "heat and pressure." They are laid together with a

little hot soldering iron put on the top of the two pieces of tin plate and then pressure is put upon them, until the heat has melted the solder sufficiently-until the pieces are fastened together— soldered by the tin coating that is on them. I think that probably the electric welding may be a perfect substitute for this process-saving the trouble and non-uniform results due to putting the variably heated blocks of hot iron on to supply the heat. I saw another very interesting thing where a piece of steel about an inch thick was put between the copper clamps constituting the poles of the machine, which were, I think, an inch and a half in diameter. The part between the clamps being in the path of the current was brought up to a bright red, the color spreading out very little further than the edge of the copper pole but extending clear through, the other adjacent parts remaining perfectly black for several seconds.

Such isolation of the heat is certainly very wonderful, and may be very useful in other processes than welding.

I was also interested in this ring business and I saw pieces bent up and welded together without the current trying to run away. Prof. Thomson, in explaining this, of course mentioned the principle that has just been explained, namely, that most of the heat went through between the poles at the joint, because the distance was so much shorter. But I want to criticise Mr. Woodbury's statement, saying that the amount of current going around the other way will be in inverse proportion to the current going through the weld according to the distance. Of course that would be the case (the cross section being uniform) if at the weld the conduction was perfect-if it was as good there as it is the rest of the way around. Then the amount of current going around and across would be exactly in inverse proportion to the distances across the weld and around the other way. But when you commence to weld, the joint is not nearly so good a conductor, because there is a break in the metal and the two surfaces are only touching each other. After a while, as the ends are pressed together and begin to weld, the metal becomes heated there, as has been explained, and is a poorer conductor, so we have two reasons for its conducting less current at that point-its being warm, and its not being in perfect contact. After a while, as the union spreads outward, the conduction becomes better by reason of more cross section. In practice, they make the pieces to be welded convex, so that they touch at the center first. This presses

and squeezes out the melted flux and all the oxides and other impurities that there are in the joint. As the bars are pushed together by pressure the metal yields a little and the impurities are squeezed out and dropped down, and finally all the surfaces to the outside are welded. By that means they know there is a perfect joint. Otherwise, if the joint were welded on the outside first the metal would contain the impurities penned up in the middle, and there would be an imperfect weld. But, speaking further about the amount of current going around a ring, when we commence there is the bad conduction due to the joint, and after it begins to heat that due to the hot metal; then, as it begins to weld in the middle, the conduction becomes better, and finally this extends clear to the outside. Then we have the condition of the metal being all one piece but hot, and even then, and until the thing is perfectly cold, the conduction is not as good there as elsewhere. Therefore it seems to me (the distance between the poles being, say one-twentieth of the whole circumference) that more than one-twentieth of the current would go around the other way. I do not know what proportion. It evidently depends upon the relative conductibility of the metal when hot and cold. If I am not right I should like to be corrected by Mr. Woodbury.

The Chairman.-Mr. Pope was asked to discuss the subject. Mr. Ralph W. Pope.-Some of the points which I wished to discuss have already been touched upon; for instance, the welding of a ring, and the shape of the abutting ends. But there have been various questions and surmises as to the practicability of the electric welding process. One practical mechanic of national reputation, and familiar with electrical work, has stated that he thought the metal liable to be burned. I wish to say in reply to this criticism, that we should not confound this process in any respect with the electric arc, as seen in the arc-light, which, as we all know, will fuse any metal. The conditions are entirely different, the current used in welding being of very low tension. The question asked in regard to the ring was a very natural one to put regarding this process. Suffice it to say that such work is actually done, and we have samples of it here to-night. I notice, however, in these samples that the links of the chains as ordinarily made have two joints. I wish to inquire of Mr. Woodbury in regard to the burr-whether it can be reduced in the manipulation, so as to bring the weld down to a size uniform with that of the bars joined. This process, as you know, has been