October 17, 1884.] 



SCIENCE. 



391 



rent, and again a discharging the battery for three or 

 four hours on this fixed resistance ; then a re-charging 

 for twenty hours; and at the end of that time the 

 battery was in good condition again, and those par- 

 ticular cells which we had found had given out before 

 did not give out again for several days; but in every 

 case, I think, where that process has been repeated 

 three or four times, it finally took about three weeks 

 for these batteries to give out. I don't remember ex- 

 actly now whether the same cells are involved, but I 

 think that out of the five or six cells, four were in- 

 volved every time. We occasionally find another cell 

 going in the same way, and I think it is a question 

 of time when all the cells will probably behave in the 

 same way: otherwise the cells which are good, and 

 which have not given any trouble, are good cells; and 

 there is no doubt about the resistance of the whole 

 battery. The electromotive force is very constant, 

 with the exception which I mentioned, and the light 

 is very satisfactory. But I am not at all sure that 

 Mr. Preece has found a solution of the light in ques- 

 tion by means of the secondary battery. It is to be 

 hoped that he has. My experience rather indicates 

 that we have not got a storage-battery, or a secondary 

 battery, which will be a practical instrument. 



Mr. Preece. — I should like to reply to some of 

 the observations that have been made; and, in the 

 first place, I should like to corroborate what has 

 been said about the percentage that the storage-bat- 

 tery would utilize in shape of light. 



Very recently Professor Dewar was kind enough to 

 make a calculation for me that brought out a very 

 interesting result, and the result was this. I wanted 

 to know the relative proportion of energy expended 

 in different modes of artificial illumination. I want- 

 ed to know how much was expended in the sperm- 

 candle to give one-candle light, and how much in a 

 gas-flame, and how much in electric lamps. The 

 result of Professor Dewar' s calculation was to show 

 that in a sperm-candle for one-candle light we ex- 

 pend ninety-seven watts. The experiment with gas 

 showed, that, for every candle given out, sixty-two 

 watts were consumed. Now, I have been experi- 

 menting with various incandescent lamps; and one 

 lamp gave a result showing that it was possible to 

 obtain one candle for two watts and half. Still with 

 arc-illumination — use of arc-lights, such as are used 

 in the street — we get a candle for each watt ex- 

 pended; and in the arc-lamp, when bereft of the 

 hideous structure put around them to destroy or 

 reduce their light, and we get all the light emitted, 

 we shall probably get a candle for each half-watt. 

 If we can get one candle for the expenditure of one 

 watt, and we do expend ninety-seven watts in the 

 process, it is clear that ninety-six watts must have 

 been wasted, as it were, in that effect. 



There is another point that arises out of this, to 

 which I want particularly to call your attention; 

 although I am afraid I am going to get into a subject 

 in which, if introduced, many of you may set me down 

 as very heterodox. My doctrine is simply this, — 

 that I believe the days of the arc-light are numbered, 

 and that all the lights in the future will be furnished 



by the incandescent lights. My reason for saying 

 that is simply this, — that, up to the present time, in 

 the incandescent lamp such as we have, the light is 

 produced by the expenditure of from four to five 

 watts per candle. Improvements, and very rapid im- 

 provements, are being made in the form of the carbon 

 filaments. We are now using lamps in England that 

 give us light with an expenditure of only two watts 

 and a half per candle. If improvement goes on at 

 this rate, I am quite certain that before another decade 

 we shall have incandescent light that will give us 

 one candle for each watt. When that is the case, 

 then the incandescent lamp will be used in place of 

 the arc. The arc requires constant personal super- 

 vision. It requires mechanism to keep it in order. 

 It only lasts a short time, whereas the incandescent 

 requires no attention whatever after it is put up; 

 and its life is very considerable indeed. In London, 

 lamps can now be obtained whose life is guaran- 

 teed to be a thousand hours. I believe on this side 

 of the water they are guaranteed to a certain extent, 

 I do not know what; but if we can get incandescent 

 lamps giving a light at the expenditure of a watt per 

 candle, and whose life will be over a hundred hours, 

 then it will be a case of good-by to the arc-lamps. 



There was one other point that I did not mention 

 in regard to my battery; and that was, I am using 

 only seventeen cells. I am using only thirty volt 

 lamps, and I use them for security, first, because it 

 gives me very few cells to keep in order; and, sec- 

 ondly, because the electromotive force is so low that 

 there is not the slightest fear of shock, and conse- 

 quently there is not the least fear of fire due to a 

 short circuit or imperfect action in any of the insu- 

 lation. In isolated houses, the lower you can reduce 

 the electromotive force, the safer it will be. 



Now, with regard to carbon : I am sorry to say that 

 I differ altogether from the view that has been 

 expressed, that carbon is likely to replace lead for 

 secondary batteries. Carbons do disintegrate with us 

 in London to a very large extent in the present form 

 of battery. I had the bi-chromate battery : the bat- 

 tery we principally use is a bi-chromate battery. In 

 that battery the carbons do not last more than twelve 

 months. They do disintegrate : they tumble to pieces, 

 they become quite soft and spongy; and it is quite 

 impossible to use any form of carbon, — moulded 

 carbon; and we are obliged to use cut carbons, as the 

 moulded carbons would not last more than two or 

 three months. While the cut carbon does not last 

 more than twelve months, the lead in the secondary 

 batteries would last more than twelve months. I 

 said that I did not agree with the remark that was 

 made, the statement that it lasted only four months: 

 I think it will last more than twelve months. I know 

 of houses where they have been in use for more than 

 twelve months, and I am quite satisfied that the 

 lead used in the one I have described will have a 

 durability of more than twelve months. I shall not 

 be the least surprised if it lasts for two or three 

 years. 



Now, I look upon the employment of the second- 

 ary battery for starting the gas-engine as simply 



