676 



SCIENCE 



[N. S. Vol. XXXIII. No. 853 



line of a heated pipe, through which the 

 rods could be automatically passed. Some 

 time was spent in trying to make a prac- 

 tical furnace out of a length of ordinary 

 iron pipe, which was so arranged as 

 to carry enough electric current to be 

 heated to the proper baking temperature. 

 Troubles here with oxidation of the iron 

 finally led to substitution of carbon pipes. 

 This resulted in a carbon tube furnace, 

 which is merely a collection of six-foot 

 carbon pipes, embedded in coke powder to 

 prevent combustion, and held at the ends 

 in water cooled copper clamps, which in- 

 troduce the electric current. By control 

 of this current the temperature could be 

 kept constant at any point desired. When 

 this was combined with a constant rate of 

 mechanical feed of the air dried rods of 

 porcelain mixture, a good product was ob- 

 tained. For the past seven years this 

 furnace has turned out all the arrester 

 rods, the number produced the last year 

 being over 100,000 units. 



In this work we were also forced to get 

 into close touch with the electroplating de- 

 partment. The rods had to be copper 

 plated at the ends, to insure good electrical 

 contact. The simple plating was not 

 enough. This introduced other problems, 

 which I will pass over, as I wish to follow 

 the line of continuous experiment brought 

 about, in part, at least, by a single in- 

 vestigation. The electric furnace consist- 

 ing of the carbon tube packed in coke was 

 a good tool for other work, and among 

 other things we heated the carbon filaments 

 for incandescent lamps in it. We were 

 actuated by a theory that the high tempera- 

 ture thus obtainable would benefit the fila- 

 ment by removal of ash ingredients, which 

 we knew the ordinary firing methods left 

 there. While these were removed, the re- 

 sults did not prove the correctness of the 

 theory, but rather the usefulness of trying 



experiments. It was found by experiment 

 that ■the graphite coat on the ordinary 

 lamp filament was so completely changed 

 as to permit of a hundred per cent, increase 

 in the lamp life or of a 20 per cent, in- 

 crease in the efficiency of the lamp for the 

 same life, so that for the past four or five 

 years a large part of the carbon lamps 

 made in this country have been of this im- 

 proved type. This is the metallized, or 

 Gem lamp. Naturally, this work started 

 a great deal of other work along the lines 

 of incandescent lamp improvement. At no 

 time has such work been stopped, but in 

 addition to it, the new lines of metallic 

 filament lamps were taken up. In fact, 

 during the past five or six years, a very 

 large proportion of our entire work has 

 been done along the line of metallic tung- 

 sten incandescent lamps. In this way we 

 have been able to keep in the van of this 

 line of manufacture. The carbon tube 

 furnace has been elaborated for other pur- 

 poses, so as to cover the action under high 

 pressures and in vacuo. Particularly in 

 the latter case a great deal of experimental 

 work has been carried out, contributing to 

 work such as that connected with rare 

 metals. In such a furnace, materials which 

 would react with gases have been studied 

 to advantage. Our experience with the 

 metallized graphite led to production of a 

 special carbon for contact surfaces in rail- 

 way signal devices, where ordinary carbon 

 was inferior, and suggested the possibility 

 of our contributing to improvements in 

 carbon motor generator brushes. On the 

 basis of our previous experience and by 

 using the usual factory methods, we became 

 acquainted with the difficulties in pro- 

 ducing carbon and graphite motor brushes 

 with the reliability and regularity de- 

 manded by the motor art. Furnace firing 

 was a prime difficulty. Here again we re- 

 sorted to special electrically heated muffles, 



