October 24, 1901] 



NA TURE 



633 



glower, as is well known, is made of a mixture of oxides 

 similar to those used in the manufacture of Welsbach 

 mantles, which are mixed to a paste and then pressed 

 through a die into threads of the desired thickness. 

 These threads are cut into convenient lengths and baked, 

 after which leading-in wires are affixed to the ends. This 

 connection, as originally made by Dr. Nernst, consisted of 

 a few turns of platinum wire twisted round the ends of the 

 glower and cemented to it by a suitable paste ; the terminal 

 connection worked out by the Westinghouse Company has 

 the platinum lead wire terminating in a small bead which 

 is actually embedded in the end of the glower, the object 

 being to prevent any shrinking of the glower spoiling the 

 contact between it and the platinum. It has often been 

 stated that if the current through a glower be altered, 

 the potential difference between its ends falls as the 

 current is increased. The curves published by Mr. 

 Wurts show that for a glower burning in air this state- 

 ment is not quite correct ; the curves only give the 

 characteristic of the glower between o'3 and o'65 ampere, 

 but they show that as the current is increased from o'3 to 

 o'5 ampere the potential difference rises also, from iSo 

 to 197 volts ; the potential difference remains constant as 

 the current is further increased up to o'55 ampere, after 

 which it begins to fall slightly. It is, however, quite 

 evident that if the glower is run anywhere near the crest 

 of this curve, as is the case in practice, a steadying series 

 resistance is essential, especially on a circuit in which the 

 voltage fluctuates, as it does on all the supply circuits 

 with which we are acquainted. 



In the above respect the glower behaves in a manner 

 comparable with the arc rather than with the carbon 

 filament. The behaviour is generally accounted for by 

 saying that the glower being an electrolytic conductor 

 becomes less resisting as its temperature .[is increased. 

 That the phenomenon is, however, more complex than 

 this is shown by the characteristics of the glower in other 

 gases than air, or in a vacuum. For a glower burning in 

 oxygenthecharacteristicappearstobealmost identical with 

 the air curve ; when nitrogen is used the curve is similar 

 in shape, but the maximum voltage occurs at a lower 

 current and is about 4 per cent, less in magnitude. The 

 characteristics for hydrogen and a vacuum show only a 

 falling curve ; there is no portion within the same current 

 limits in which the potential difference rises wMth the 

 current. These results seem to indicate that some 

 chemical changes are going on between the glower and 

 its surrounding envelope of gas which have an impor- 

 tant influence on its behaviour and very possibly on its 

 efficiency and life. There can be little doubt that when 

 the Xernst lamp is easily available it will open up a field 

 of research as interesting and possibly as fruitful as that 

 afforded by the electric arc. 



The work of M. Blondel and Mrs. Ayrton has made 

 us familiar with the physical necessity of a resistance in 

 series with the arc on account of its falling characteristic. 

 Such a resistance is still more necessary with the Xernst 

 glower — since a smaller overload of current is more 

 dangerous. With an arc a comparatively small " dead " 

 resistance in series is sufficient, that is to say a resist- 

 ance which remains of practically constant value 

 throughout the working range of current and does not 

 heat sufficiently with a small overload to alter 

 appreciably. But with the glower it is easy to see, 

 from the curves published by Mr. Wurts, that if a 

 dead resistance were used it would have to Ise so large 

 for safety as to greatly diininish the efficiency of the 

 complete lamp. A resistance has therefore had to be 

 sought which will increase very rapidly w'ith increase 

 of current, and the solution of the difficulty has been 

 found in using fine iron wire nearly at a red heat, the 

 wire being enclosed in a small sealed bulb filled with 

 hydrogen to prevent its oxidation. This " ballast," as it 

 is named by the Westinghouse Company, has a remark- 



NO. 1669, VOL. 64] 



ably high corrective power ; with a rise in current of 

 15 per cent, the resistance increases about 180 per cent., 

 the potential difference between its terminals increasing 

 200 per cent. As a consequence with such a ballast in 

 series with the glower, a rise in the supply pressure does 

 not produce a serious increase in the current. It follows 

 also that the Nernst lamp (consisting of glower and 

 ballast) cannot be overrun in the same way as an 

 incandescent lamp in order to get it to give more than 

 its rated candle-power or to work at a higher efficiency. 

 With an increase of the supply pressure of 5 per cent, 

 the candle-power of an incandescent lamp increases 

 40 per cent., whereas that of the Nernst lamp only rises 

 about 7 per cent., the corresponding increase in efficiency 

 being 26 per cent, and 2 per cent, respectively. It is 

 obvious that as this result is attained by running the 

 iron wire nearly at a red heat, the resistance of the 

 ballast when cold is considerably lower than its correct 

 working resistance, and there is therefore a tendency for 

 the glower to take too much current when it first lights 

 up. The ballast has for this reason to be constructed in 

 such a way that the iron wire shall very quickly assume 

 its full resistance, a requisite satisfied by the free 

 suspension of the wire in the hydrogen bulb. 



Mr. Wurts does not give any curves showing the 

 candle-power and efficiency during the life of the glower, 

 a thing to be regretted, as it is by the performance of the 

 lamp in this respect that it must ultimately stand or fall. 

 It is, however, stated that the average life of 220-voIt 

 glowers on a by no means over steady alternating circuit 

 was found to be from Soo to 900 hours, and that neither 

 the efficiency nor the candle-power fell off much during 

 this life. The life of the i lo-volt glowers is said not to 

 be so good, and the average life of the glowers on direct 

 current is given as only about 250 to 300 hours. It 

 should not be assumed that this is an inherent defect, as 

 the Westinghouse Company have devoted their attention 

 mainly to alternating current, and since the German 

 company make lamps for either type of supply it seems 

 evident that the difficulty can be overcome. 



W> have done no more than indicate a few of the 

 many interesting considerations raised by Mr. Wurts' 

 paper ; it contains in addition descriptions of the heater, 

 of the cut-out used for breaking the heater circuit when 

 the glower lights, and of the complete lamp as now con- 

 structed in .America. Suffice it to say that the lamps are 

 made in candle-powers from 50 to 2000, the higher candle- 

 powers being obtained, not by the use of a larger glower, 

 but by mounting a number of glowers in parallel in the 

 same lamp. Smaller candle-powers than 50 have not 

 been made, as it has been deemed advisable to design 

 the lamp to compete with the arc rather than with the 

 incandescent lamp. In conclusion, we must agree with 

 Mr. Steinmetz, the president of the American Institute, 

 that the lamp should riot be compared too severely with 

 existing standardised lamps since, as it is just fresh from 

 the laboratory, it is to its promise rather than to its per- 

 formance that one should look. And there can be little 

 question but that its promise is e.xceptionally good. 



FLORAS OF THE PAST} 



THIS " extract " from the twentieth annual report of 

 the U.S. Geological Survey forms a bulky volume 

 of more than 200 pages and 1 50 plates. Prof. Lester Ward 

 tells us in the introductory remarks that the aim of the 

 memoir is " to give a succinct account of the progress thus 

 far made in the direction of developing the Mesozoic 

 floras of the United States." The present instalment 

 deals with the vegetation of the Triassic and Jurassic 



1 ".Status of the Mesozoic Floras of the United States." First Paper : 

 The Older Mesozoic. By Lester F. Ward, with the collaboration of W. M. 

 Fontaine, A. Warner and F. H. Knowlton. Pp. 211-430 -(- Plales x.\i- 

 clx.\ix. (Washington : Government Printing Ofiice, 1900.) 



