292 



NA TURE 



[May 6, 1909 



first a yellow and then a blue tint like steel, but when 

 the heating- is continued it burns to pentoxide. The black 

 powder and thin wires can even be lighted by applying a 

 match to them, as the experiment shows. 



Its melting point in vacuo lies between 2250° C. and 

 2300° C, which makes it particularly suitable for elec- 

 trodes in vacuum tubes, especially as it does not dis- 

 integrate. For example, it is extensively used in Rontgen 

 tubes. Its specihc weight is i6-6. 



Turning now to the electrical qualities of tantalum, its 

 specific resistance was stated by Dr. von Bolton in 1905 

 to be, on the average, 0165, with a temperature coefficient 

 of 3 per cent, between 0° and ioq° Celsius. 



Further experiments conducted by Dr. Pirani in the 

 laboratory of Siemens and Halske revealed the fact that 

 wires of various thicknesses varied in their specific resist- 

 ance from 0173 to o-i88, but after they had been heated 

 to 1900° C. in a high vacuum for from 100 to 200 

 hours, they all possessed the same specific resistance, viz. 

 0-146, and their temperature coefficient between 0° and 

 100° C. had risen to 0-33 per cent. 



As the temperature of a tantalum filament, when con- 

 suming i-s watt per candle-power, is about 1850° C, and 

 its resistance about six times its resistance at 100° C, 

 the temperature coefficient between 100° C. and 1850° C. 

 may be taken, on the average, as 029 per cent. 



No doubt the difference between these results is caused 

 by alterations in the structure of the wires during their 

 manufacture, and the heating in vacuo served a similar 

 purpose to the annealing of steel, so that Dr. Pirani 's 

 results published in 1907 may be taken as standards. 



.^t present the most important industrial application of 

 tantalum is its use for filaments of incandescent lamps, 

 which may be said to date from July, 1903, when Dr. 

 Feuerlein succeeded in producing a tantalum wire 

 one-twentieth of a millimetre in diameter. Of this wire 

 he made a glow lamp with a filament 54 mm. long, using 

 a current of 9 volts 0-58 ampere, and giving a light of 

 35 candles (Hefner), at the rate of 15 watts per candle- 

 power. 



A simple calculation shows that for a current of no volts 

 660 mm. of the same wire would be required, giving at 

 the same rate of consumption of energy a light of 43 

 candles. 



In carbon lamps, for 220 volts the length of filament 

 is only 400 mm., and the filaments remain hard until they 

 disintegrate. Tantalum filaments, like other metallic 

 filaments, soften, however, to such a degree that they 

 cannot be used in the same shape as carbon filaments. 



After trying various methods of housing the long Ta 

 filament in a glass bulb of approximately the same 

 dimensions as the carbon glow lamps, the present form 

 was arrived at during the year 1904. In this lamp, vi/hich 

 was adopted as standard, the length of the filament was 

 650 mm., its diameter 005 mm., and its weight 0022 gram, 

 so that about 45,000 of these lamps contain i kg. of Ta. 



Since then these dimensions have been modified to a 

 certain extent ; for instance, the diameter of the filament 

 is now only 0-03 mm., but the external shape has not been 

 altered. 



It was soon found that after burning a short time the 

 filament underwent certain structural changes and lost 

 its great tensile strength. Examination under a micro- 

 scope revealed the fact that in about 1000 hours the 

 smooth, cylindrical filament shows signs of capillary con- 

 traction, as if the cylinder was going to break up into 

 a series of drops, and the surface, from being dull, com- 

 mences to glitter. This contraction of the filament after 

 being heated is readily recognised by comparing a new 

 lamp with an old one. On the stars of the new lamp the 

 filament hangs loosely, while in the old lamp the filament 

 is evidently in tension. 



The characteristic difference between carbon filaments 

 and tantalum filaments is shown by a diagram represent- 

 ing the influence of temperature on the electric resistance 

 of the two filaments in proportion to each other. 



In order to have the differences at once shown in per 

 cents., the normal pressure and the normal resistance of 

 both filaments, when giving the light of i candle for 

 1-5 watts, is marked as 100, and it is immediately seen that 



KO. 2062, VOL. 80] 



the resistance of Ta alters directly, and that of carbon • 

 inversely, as the temperature. Owing to this quality a 

 Ta filament is belter able to resist overheating than a 

 carbon filament, as the following experiment shows, where 

 two lamps, one Ta and one C, burning normally at 

 no volts with 1-5 watts per candle-power, are gradually 

 exposed to higher voltages. The C lamp breaks, while 

 the Ta lamp stands up to 200 volts, the highest voltage 

 available here. Of course, its useful life will be shorter 

 than at its normal voltage. 



As stated at the beginning of the discourse, the primary 

 object of all the research was to find a filament more 

 economical in the consumption of electrical energy than 

 the C filament, and the following experiments will show 

 that the Ta filament is in this respect a great improve- 

 ment on the C filament. To begin with, a comparison can 

 be made by burning a Ta and a C lamp under water, each 

 being immersed in a vessel containing the same quantity 

 of water. Owing to the C lamp requiring more energy 

 to give the same light as the Ta lamp, the temperature 

 of the water in the C vessel rises quicker than in the 

 other vessel. Another way of showing the difference is 

 by measuring the current taken by each of the two lamps 

 when giving approximately the same light, or by send- 

 ing the same current through both lamps in series and 

 noting the difference in candle-power. 



In conclusion, two interesting qualities of Ta should 

 be noted. The first is that, when a Ta filament is heated 

 in a high vacuum, it will expel any oxygen that has com- 

 bined with it. It is possible to detect whether a filament 

 contains any oxide by very gradually heating it up, when 

 the parts containing oxide will appear brighter than those 

 consisting of pure Ta, owing to the greater electrical 

 resistance of the oxide. 



These lamps have been purposely exposed to the air 

 while they were being exhausted, and have become 

 " spottv ** in consequence, but if they are raised a little 

 above their proper voltage and left burning for a few 

 minutes their filaments become quite uniform by the ex- 

 pulsion of the oxygen. 



The second quality is that Ta will act as a rectifier when 

 used in an electrolyte, that is to say, it will allow the 

 passage of the positive current only in one direction. In the 

 apparatus shown the positive current passes through the 

 lamp to a Ta anode, thence to a Pt kathode, but in a very 

 short time the Ta anode covers itself with a film of oxide 

 which stops the current. When the current is reversed 

 the lamp lights again, and continues to burn. When an 

 alternate current is connected to the lamp it will also 

 continue to burn, but with diminished brilliancy. 



All these experiments are intended to show the remark- 

 able qualities of this material, and when they are fully 

 appreciated and its limitations are properly understood 

 there appears to be a great field open to tantalum and its 

 industrial applications. 



CONFERENCE ON ROADS. 

 A CONFERENCE arranged by the County Council 

 •'^ Association, in conjunction with the Association of 

 Municipal Corporations, the Urban and Rural District 

 Councils' Association, the Association of Municipal and 

 County Engineers, and the County Surveyors' Association 

 was held in London last week, and lasted over three 

 days. The meetings were divided into three sections, 

 which met at the Institution of Civil Engineers, the 

 Mechanical Engim.-ers, and the Surveyors' Institution. 



This conference was very largely attended by borough 

 and county engineers, chairmen and members of the Roads 

 and Bridges Committee of county councils, and others 

 interested in automobiles. 



Following on the International Road Congress held at 

 Paris last year, this gathering together of those responsible 

 for and interested in the management of the highways of 

 this country shows the increasing importance of road 

 traffic and of the interest taken in the condition of our 

 highways. 



Forty papers were contributed for reading and dis- 

 cussion, the subjects dealt with relating to the construction 

 and maintenance of roads ; motors and traction engines. 



