As far as Dr. Sharp knew, no 110-volt lamps have been produced as yet for lower candle-powers than 25 and no lamps have been produced for higher voltages than 220. The fineness of the 110-volt, 25-cp filament is such that it would seem to be difficult to produce such a lamp as a regular commercial article. The 220-volt lamps are probably only experimental as yet. The properties of the tungsten filaments are such that it would seem to lend itself very readily to the production of most excellent lamps. for street lighting by the series incandescent system. Lamps for 110 volts are likely to have, when commercially produced, a watt consumption of 50 watts or more. If tungsten lamps are to be made for small candle-powers, such as are commonly employed in domestic lighting, they would probably need to be made for 50 volts or under, and consequently either burned in series or connected to low-voltage mains.

One of the chief disadvantages of the tungsten lamp lies in the extreme fragility of its filament. Blows or shocks given to the lamp are quite likely to cause a rupture of the filament. A ruptured filament may, however, mend itself by the parts welding together once more, but where the filament has become welded it is quite likely to break again.

The feature which differentiates the electrical behavior of the newer lamps from the ordinary carbon lamp is their positive temperature coefficient. The effect of this is to give the lamps. considerable inherent regulation. That is, the change of current through the lamp is no longer proportional to the change in voltage, but is smaller proportionally than this change. Consequently, in candle-power and watts per candle, these lamps undergo smaller changes with the change in the line voltage than is the case with the carbon-filament lamp. This is shown in the accompanying table, which gives the change in candle-power and the watts per candle of carbon, metallized, tantalum and tungsten lamps with 5 per cent rise in the voltage.

The quality of the lamps here pointed out is a very valuable one, since it must have two important results: (1) The light of the lamps is less affected by bad regulation of the circuit. This means that with a given degree of regulation of the voltage on the circuit the service must be more satisfactory to the user, and has a direct bearing on the amount of copper required in feeders. (2) The life of these lamps is probably less affected by the momentary or even continued application of excessive voltages.

As to the development of the tungsten lamp abroad, it was stated last month by the importer of one type that the German factories making it are turning out now about 5000 lamps per day and the Hungarian factories about the same quantity. These lamps are 20-cp and 30-cp for 100 to 125 volts, and 50-cp for 200 and 250 volts, the latter with only 5 filaments. Those brought into this country thus far are of the 40-cp type, with an average life of 1000 hours and a consumption of I watt per candle.

Since the metallic-filament lamps taken up in this country have all been practically of foreign origin, it deserves mention that Mr. J. Allen Heany, of York, Pa., has long been working in this field, and has obtained very definite results, following original methods. A very recent patent describes his process of manufacture of a homogeneous titanium incandescent lamp filament. The claims of the patent are on a filament made of pure metallic titanium "of high fusing point and electrically conductive, which is stable at an efficiency at which a carbon filament, or a titanium carbide filament, will rapidly disintegrate, the lightemitting properties of the filament being due to homogeneous nature of the titanium;" and on the process of forming the filament "which consists exclusively of a conducting titanium compound by passing therethrough an electric current in a vacuum, thereby illuminating the substance with which the titanium was combined with and by means of said electric current shrinking and sintering the remaining pure metal into a dense homogeneous shiny filament." The specification states that the invention relates to the manufacture of luminants for electric lamps made of very pure refractory metals, such as tungsten, titanium, zirconium, and so forth, or alloys of two or more such metals, or of pure chromium, molybdenum, thorium, manganese, or alloys of such metals.

These luminants are ductile and capable of standing a much higher temperature than any form of carbon or the carbides of such metals, and are more efficient than any hitherto known metallic filaments. They can withstand a much higher temperature and also convert the heat into light waves, and thereby have the properties of selective radiation. These metals, or osmium, cerium, niobium, tantalum and vanadium, or boron and silicon, are utilized, either singly or mixed in a powdered form, as some of these powdered metals can be obtained by known processes more or less pure; or the pure oxides, hydrides, nitrides, or metals in a very fine powder or in a colloidal state of the oxide or collodial suspension of the pure metal can be employed, and with the dry powder a lubricant or binder used, such as water or paraffin, to form or shape them. They are then baked in an oven to drive out the paraffin or water, and in case the oxide is used the filament is reduced in pure hydrogen by external heat to the metal. The preliminary baking does not oxidize the metal, but merely strengthens it, and during this operation the binder is dissipated, leaving the filament formed of a refractory substance, strong, durable, and of good conductivity. The body is now heated by an electric current in a vacuum to drive out the hydrogen and to sinter or alloy the particles. If the filament is too large or rough, it may be rolled or drawn to the desired shape. Dr. Steinmetz, it may be noted, in a discussion before the American Institute of Electrical Engineers, has expressed an opinion rather derogatory to the value of the property of ductility in tungsten, in connection with incandescent lamps. Heany tungsten lamps tested at the United States Bureau of Standards are reported as consuming 0.95 watt per candle.

There does not appear to be anything strikingly new to report as to osmium-filament lamps, but one or two developments have occurred in connection with zirconium. A new lamp has been made in Germany by the Zirconium company for voltages to 200, with an approximate power consumption of one watt per hefner-candle (0.909 candle). For voltages of 110 the lamp is made for 35 hefner-candles or more, for 220 volts for 70 hefnercandles only. No exact information is given on the composition of the filament. In the prospectus of the company it is claimed

that the life is from 500 to 1000 hours without appreciable diminution of the candle-power during this time. It is also claimed that the lamps do not blacken. The following results were obtained in practice with the new lamps. An experiment with 110-volt, 56-cp lamps gave bad results. Most of the lamps were destroyed in less than 150 hours, and some became quite black during this time. A somewhat better result was obtained with the 110-volt, 35-cp lamp. Three lamps were selected from one shipment. Of these the first had a life of 350 hours, the second 450 hours, no appreciable diminution in the candle-power being observed during this time. The third lamp alone was still in service when reported, having been operated for 600 hours, but being already considerably blackened. The average life of 500 hours claimed seems too high.

In this connection further note was made by the European journals in January of the "zircon-wolfram" filament invente 1 by Zernig, and suitable for a range of from 200 to 220 volts. The filament was made originally from a mixture of zircon and carbon. This gave a high-efficiency filament, but only suitable for low pressures. Subsequently tungsten was introduced. As to the exact process used, it is difficult to obtain any exact information, but it appears that it depends upon utilizing hydrogenous combinations of metals. The material that is obtained is in such a form that the filaments can be made by a process of squirting. The introduction of tungsten into this particular lamp has enabled an extraordinarily rapid advance to be made in the pressure at which it will run. Thus, although the pressure was only 37 volts rather more than a year ago, this pressure has now been raised to 220 volts. Lamps are at present made for 16 candles on 100 volts, or 32 candles on 200 volts. The cost of manufacture is said to exceed that of the carbon-filament lamp by not more than 2 cents per lamp. The lamp is claimed to run equally as well on alternating as on direct current. It is not fragile, and the consumption is given as about 1.2 watts per candle. Discussing this lamp before the British Institution of Electrical Engineers, Mr. James Swinburne remarked recently that he considered it very important as the first 200-volt metallic-filament lamp. It had been stated that the lamp is made of zirconium and tungsten, but he did not agree that it was an alloy at all, as alloys have a

low melting point. It is more probably a zirconide. The lamp, he said, may possibly be the first of a new series in which such compounds may play an important part.

On this point, Boehm may be cited. In a paper on the chemistry of new electric incandescent lamps, he refers to the zirconium lamp, which should really be called a zirconium-carbide lamp, since the filament does not consist of metallic zirconium, but of its carbide. The first lamps of this kind consumed about 2 watts per hefner-candle, or more. Recently an improved lamp of this type has been made for which a consumption of I watt per hefner-candle is claimed. This is a carbide lamp in which "the content of carbon is reduced by suitable additions and methods." As suitable additions mention is made of rutheniumi and tungsten. Such filaments have a much higher melting point. and can be operated with more watts per unit of surface and with a reasonably long life. According to Boje, these lamps may be operated at as low a consumption as 0.3 watt per candle. With a consumption of o.6 watt per candle their life is more than 120 hours, and with 1 watt per candle it is more than 1000 hours; during the first 500 hours the candle-power is almost constant. The thickness of the filament is 0.6 millimetre, and its length per volt is about 5 millimetres, so that several filaments must be connected in series in order to construct lamps for 110 and 220 volts.

The only other metallic-filament lamp to note from the standpoint of practice is the iridium, which, however, like the osmium, is suitable only for low voltages. Moreover, it can not compete with carbon, and the others, as the supply of iridium is limited. A storage-battery company in Germany has undertaken its manufacture in connection with battery plants.

DEVELOPMENTS OF ARC AND OTHER LIGHTING

Not less than the incandescent branch of the art, the field of arc lightng has been alive with advances and improvements during 1906-7. There have been developments with the enclosed arc, the flame arc has become a more pronounced success, and even the old straight-carbon arc has shown a vitality and persistence with which it had not been credited. Probably most interest has attached to the flame arc, now becoming so familiar an