The positive, or anode, used with these metallic-oxide negatives is generally a metal and is consumed much more slowly than the negative. This is contrary to what would be expected, judging by the action of carbon electrodes. There are a number of advantages possessed by the metallic arc over the carbon arc. In the first place, the efficiency is much better; that is, a metallic arc lamp operating on a 4-ampere current with approximately 65 to 70 volts at the arc will give a light equal or superior to that of a 6.6-ampere, 75-volt, directcurrent, enclosed-carbon arc lamp. The distribution of light is far better. This is owing to the fact that in the enclosed-carbon arc practically all the light comes from the crater on the flat under-surface of the upper electrode, most of it being thrown down and not serving to illuminate the street between lamps. The light from the carbon arc itself is weak and of a blue color. This is very pronounced at times, especially if the flat under-surface of the upper electrode is somewhat inclined, thus hiding the crater. In the case of the metallicoxide electrodes, the arc is itself the source of light, practically none coming from the crater, except by reflection. The metallic arc is much like a candle flame, having its luminous and nonluminous zones. The light is brightest near that end of the arc which is next to the negative electrode, and comes from a hollow cone-shaped mantle of volatilized oxide of titaniumı rendered incandescent by the heat of the arc, just as in the candle flame the light comes from a hollow cone-shaped mantle of carbon particles made white-hot by the heat of the flame. The voltage required to maintain a metallic arc is less than that of an enclosed-carbon arc. It is a familiar fact that an enclosed-carbon lamp will not burn properly with the arc voltage down to 65, while a metallic arc will burn well at less than 55. Metallic arcs are adjusted to burn at from 65 volts to 75 volts in different cases, while the carbon arcs are all set at 80. This is a very evident advantage in favor of the metallic arc, as more lamps may be put on a circuit without raising the voltage on the line. The life of carbon electrodes, as a rule, is not over 150 hours, while the metallic-oxide electrodes can go considerably longer. The uniform white color of the metallic arc is in marked contrast to the changeable blue and white of the enclosed-carbon arc. As the metallic-oxide electrodes are not burned "enclosed," there is no inner globe required on the lamp. While it looked easy to secure all of these advantages, many difficulties appeared, but they have now practically all been overcome. In the first experiments the electrodes were trimmed with the anode or positive above and the negative or metallic-oxide electrode below, just as carbon lamps are trimmed, but a number of troubles presented themselves. First-The bright portion of the arc was near the surface of the lower electrode, which cast a large shadow. Second-The light reflected from the brilliant surface of FIG. 1-A-METALLIC ARC WITH NEGATIVE BELOW C-METALLIC ARC WITH NEGATIVE ABOVE the fused slag on the lower electrode was thrown upward and could only be partly saved by using a reflector. Third-An under-feed mechanism was seen to be necessary, as, contrary to the action of carbon electrodes, the negative metallic-oxide electrode is the more rapidly consumed. Fourth-Only a comparatively short metallic-oxide electrode could be used, as a long one would necessitate the use of an unwieldy long glass globe. This would limit the life and could only be met by adopting a negative electrode of large diameter, which it is evident would be undesirable. Fifth-A particularly undesirable feature was the gathering of a large amount of reddish soot that would collect in spongy masses around the electrodes, obscuring the light. This had to be removed by some mechanical means, such as scraping or shaking it off, and some receptacle other than the glass globe had to be provided to catch it. Sixth-The negative or metallic electrode was seen to burn to a blunt taper point, causing the arc to be very unsteady, as it tended to leave the end and run up the side in the manner of the carbon arc when flaming. As noted above, the bright portion of the metallic arc is located near the surface of the negative electrode, and it was seen to be very desirable to burn the electrodes with the negative above, thus getting the bright portion of the arc in such a position that the shadow thrown down would be less, and that the light reflected from the brilliant surface of the fused pool of slag on Soot B FIG 2-A-SHOWING HOW THe Ends of ELECTRODES WASTED AWAY TO the negative electrode would be thrown down and utilized instead of being thrown upward and wasted. The other advantages, noted above, possessed by the carbon lamp would be retained if this inverted position of the electrodes could be made practical. The first attempt to burn the metallic-oxide electrode above and the metallic electrode below showed that there were serious obstacles to be overcome before it could become a practical method. In the first place, the electrode would not keep a square end, but would waste away on one side, and the arc would run up this bevel, or slope, giving a very unsteady light. In the second place, the volatilized oxides of iron, chromium, titanium, and so forth, would condense on the sides of the electrodes and hang down as a fringe or curtain, hiding the light. The first means taken to overcome these troubles was the introduction of a rotating draft of air around the arc. This had the effect of forcing the arc to hold to a central position, stopped the crooked burning, and steadied the light, but did not take care of the fumes. Attempts to blow the fumes away sidewise gave only partial success. Finally, a current of air was directed down around the arc, and this gave excellent results. The electrode burned perfectly square, and the clean layer of air prevented any gathering of fumes. This was a very marked advance, as this did away with any need for a mechanical scraper or To chimney FIG. 3-THE DOWN-DRAFT OF AIR FORCES THE SOOT TO TAKE THE PATH SHOWN BY THE DOTTED LINE, EFFECTUALLY KEEPING IT FROM TOUCHING OR DEPOSITING ON ANYTHING shaker, the soot practically all passing out of the chimney and not requiring to be caught in any receptacle, the globe remaining clean. When burning metallic-oxide electrodes with the metallicoxide stick below, copper was used as an anode with fair results. On reversing the position of the electrodes, it was found that the new conditions made it possible to improve on the action of a pure copper anode, and a number of changes were accordingly made. In the first place, if the arc plays for a time on pure copper, it will oxidize the surface. This oxide will fuse to a slag that becomes an insulator when cold, and on starting a cold lamp it is necessary to strike the electrodes together hard enough. to break through the slag. To strike such a hard blow is undesirable, as, if it is done while the lamp is burning-for example, when feeding-it is liable to spatter the fused slag out on to the glass globe. A simple remedy for this consisted in using an anode containing metals or alloys whose oxides, when fused together, would make a slag that is a good conductor when cold. The steadiness of the light largely depends on the composition of the slag, its uniformity and temperature. The anode surface is at all times covered with this slag, which slowly dissolves the metal and is itself slowly volatilized. If the arc plays on bare metal it consumes it rapidly, and it was found desirable to secure this slag cover from being knocked off. This was accomplished by providing a rough surface for it to cling to, and by running the entire anode tip very hot. A characteristic property of the metallic arc has been a very FIG. 4-PATH TAKEN BY AIR CURRENTS IN LAMP noticeable dying-down or dimming of the light, which would occur at irregular intervals, especially after the electrodes had burned for 20 hours or more. These dim spells would last from a few seconds to two or three minutes, when the normal brilliancy would return. This is explained as follows: In the metallic arc, the brilliancy is largely due to the presence of volatilized oxide of titanium, and anything that interferes with the uniform evolution of vapors of titanium will cause the light to dim-for example, the presence of a high percentage of highly infusible oxide of chromium. This oxide of chromium is volatilized at a slower rate than the oxides of titanium and iron, and after the electrode has been burning some 20 hours the slag on the end of the cathode has become very rich in oxide of chromium, which forms a film on the surface of the fused pool of oxides. |