l62 



NATURE 



[December 18, 1902 



one hand by the hysteresis loss in iron, which prevents our using 

 higher inductions in armatures, and low permeability, which 

 limits our field and armature tooth inductions. It does not seem 

 likely that we shall now find iron much better in either respect. 

 Nor are we likely to find a better available conductor than pure 

 copper. As insulator we have mica. It looks, therefore, as if 

 we were within sight of our limits in dynamo and motor designs. 



Secondary Batteries. 

 The secondary battery in central station work has been used 

 as a store to equalise the load, and to reduce the running plant 

 at the times of heavy load. Owing to the high full-load station 

 pressure with feeder systems, the station battery is generally for 

 use at light loads only. But the secondary battery has for a 

 long time been on the border of success for traction work, both 

 on tramways and on the road, and a further improvement in 

 batteries may be expected to produce very great changes in im- 

 portant branches of engineering. 



The first question asked is, Why do we stick to lead ? The 

 answer is that the case is very special and other things will not 

 do. We are practically limited to lead, at any rate in acid cells. 

 Take first the plate that oxidises on discharge. It should not 

 dissolve in the electrolyte, as if it does the deposition and solu- 

 tion will be uneven, and the plate will grow trees and come to 

 grief. This puts zinc out of court, unless some electrolyte is 

 used which gives some insoluble salt of zinc, which does not 

 attack zinc on open circuit, and gives a good electromotive 

 force with it. Iron is out of court for the same reason ; there 

 is no suitable electrolyte. The strong organic acids such as tri- 

 chloracetic or oxalic are apt to have their positive radicles 

 split up by electrolysis, even if a strongly positive metal can be 

 found with an insoluble salt. Lead is thus the only metal practi- 

 cally available in an acid electrolyte. Silver in hydrochloric 

 acid would give no pressure, and the acid would be decomposed 

 at the anode. On the other plate we need an insoluble de- 

 polariser, else a two-fluid cell must be used, involving a porous 

 diaphragm, diffusion and impracticability. Not only must the 

 depolariser be insoluble, but it must be converted into an in- 

 soluble body on discharge. The coating must be a conductor in 

 one state or the other, or there will be no proper contact. In 

 the lead cell, there is always enough peroxide and metallic lead 

 in the coatings to secure electrical contact though the discharge 

 product is an insulator. The depolarising coating must be con- 

 nected to a conducting plate which is not attacked by local 

 action. Lead and silver are the only available metals, and 

 sulphuric, and perhaps phosphoric, the only acids, for the nitrate 

 of lead is soluble and hydrochloric acid is decomposed by lead 

 peroxide. Lead is protected by its coating of sulphate, or per- 

 oxide as the case may be. It thus seems as if we were limited 

 almost absolutely to lead and sulphuric acid. It is wonderful 

 that we 'have the lead cell at all. We owe it to the chance ob- 

 servation of Plante. The theory was not understood for a long 

 time. For many years it was thought that the pressure was due 

 to the PbO and Pb changing into PbO. The acid was merely 

 put in to make the electrolyte conduct, and sulphuric acid was 

 used because people used it in gas voltameters, and they never 

 thought that it ought to be as strong as practicable to give the 

 pressure and output. The formation of lead sulphate was regarded 

 as a difficulty to be overcome. 



In the lead cell we want lightness, large capacity, cheapness, 

 rapid discharge, efficiency and mechanical strength, and dura- 

 bility. These qualities are mostly antagonistic. Large capacity 

 means rapid deterioration. Mechanical strength means weight. 

 It is thus no use testing a cell for capacity without testing the 

 efficiency and durability too, and so on. Published battery 

 reports are often misleading, because they omit essential 

 information. 



Cables. 

 The conductor itself can hardly be improved, but there is 

 great room for improvement in the insulation. It is largely the 

 insulation of the cables that limits our pressures, and therefore 

 our distances of transmission. For iooo kilowatt cables, the 

 cost is about a minimum for 8000 volts ; above that, the cost of 

 insulation increases faster than the cost of copper falls. It is 

 exceedingly unlikely we have reached the limit in insulation. 

 There is no branch of electrical engineering so important as 

 cable making. Cables form a large portion of the capital outlay 

 in large systems. Vet there is no branch of the industry which 

 is run on less scientific lines. The days of secret mixtures 

 known only to the workman who makes them may be passing 



NO. 1/29, VOL. 67] 



away ; but even now the whole art of cable-making is a question 

 of trial and error, with a good deal of the last component. 

 Engineers do not know now whether rubber is better than 

 paper, nor can they tell what any particular make of cable will 

 be like after ten years' use. 



Light. 



Our chief work, until lately, has been producing light. Here 

 the inefficiency and waste is prodigious, and though it is mostly 

 unavoidable, there is still great room for improvement. We 

 take great care over our stations, watching every penny from 

 the coal shovel or mechanical stoker to the station meter. We 

 quarrel over l per cent, in the generators. When we get to 

 the mains we care less, and once we have got to the consumers' 

 meters we care nothing at all. 



Practically all light is wanted for use by the human eye. The 

 human eye is exceedingly sensitive ; it is calculated to see a 

 distant star when receiving io -8 ergs per second, so that one 

 watt would enable, say, five thousand billion people to see stars 

 with both eyes, but it would have to be used economically. In 

 reading a book, the eye would need much more than this ; and 

 then, as the book radiates light in half of all directions, only a 

 little is used by the eye, so even if all the light from a source 

 were concentrated on a book, there is enormous waste by useless 

 radiation from the book. But the source of light does not illu- 

 minate only the book ; the book probably subtends a small solid 

 angle, so we have another source of waste. The eyes reading a 

 book in a fairly good light want something of the order of two 

 ergs per second, so that a watt would only work the optic nerves 

 of, say, the inhabitants of London. But the book, say 200 

 square centimetres, would need about 3000 ergs a second to 

 illuminate it. A candle, which gives a light of 4T, radiates 

 about 02 watt, or five candles a watt ; that is to say, at an 

 efficiency of unity, we would get five candle-power or 20 units 

 of light per watt. The efficiency of a glow-lamp is only about 

 025 candle-power per watt, or ox>5, so there is room for im- 

 provement. The first thing, naturally, is to see what limits 

 there are in the way of increased efficiency. The obvious goal 

 is direct production of "light without heat," by which is meant 

 producing only the rays of wave-lengths which affect the eye. 



There is no thermodynamical reason why electrical energy 

 should not be converted directly into radiation of any wave- 

 length without loss ; I do not know if there is any molecular 

 impossibility, but apparently our limits are practical — that is to 

 say, it may be done, but we have not yet hit on the way of doing 

 it. The vacuum tube appears to be a means of converting 

 electric power direct into radiation. The Cooper- Hewitt lamp, 

 for instance, gives an efficiency of about three candles per watt, 

 or something like o -6. All these figures as to light are a little 

 vague. Unfortunately, the light is of a very bad colour. It is 

 very actinic, but the wave-lengths are too small. One method 

 is to degrade the light by making it act on silk dyed with matters 

 which lower the radiation to a redder colour by fluorescence. 



The Arc Light. 

 The arc has been very fully studied in some directions and 

 not in others. Most makers of arc lamps seem to devote their 

 whole attention to the mechanism, and look upon the arc 

 merely as a hot gap that has to be preserved by suitable 

 apparatus. Many lamp makers, on the other hand, have records 

 of exhaustive experiments on the relations of the pressure, 

 current and light with different carbons ; but they are very 

 seldom published. On the other hand, an enormous amount of 

 laborious experiment on such points as these is available, and 

 on the back electromotive force of the arc. The physics of the 

 arc, an exceedingly difficult branch of study, has not received 

 much systematic attention yet. The crater of an arc is, no 

 doubt, heated to the point of volatilisation of carbon at the 

 pressure of the air. If other gases get at the crater, the vaporis- 

 ation temperature would be less. (There is a small increase of 

 pressure which I suggest is due to the electromagnetic effect of a 

 current localised in a conducting fluid. This may be neglected.) 

 The crater may be rough, as carbon, though it softens, does not 

 melt before volatilising, and it may be merely speckled with 

 points at its volatilising temperature, so that its brightness is 

 not uniform. But there are so many anomalies about the arc 

 that one cannot say anything definite with safety. For instance, 

 if the temperature is limited by the vaporisation of carbon, what 

 must be the specific heat of vaporisation of carbon ? Where 

 does the vapour go, and what happens to it in an enclosed 

 lamp? In condensing into smoke.it should give light of the 



