fluid oil which adheres very firmly to the bearings makes the best lubricant, except in cases of very heavy pressures on the bearings, where the film of oil is too thin to keep the metal surfaces separated from each other. The peculiar adhesion of some kinds of oil to metals is well illustrated by lard oil adhering to the cutting edges of machine tools better than other oils. I presume in these experiments that the spindle band tensions were made constant; that Mr. GOODALE took all possible care that there should not be any variation in the spindle bands beyond what is inevitably due to change in the humidity of the air. I would like to ask if there was any change in the oil used in other parts of the spinning frames except the spindles. Mr. GOODALE. The oil on other parts was kept constant, every care was taken with the bands, and the humidity was kept up.

A MEMBER. I would like to have Mr. GOODALE tell us how we are going to guard against variations in the quality of the oil. Every little while I find our machinery running hard, and that we lose more power on some days than on others, for which there is of course some reason. Although we buy our oil of the same brand, from the same parties, it seems to come to us of various qualities; and if there is any simple method which will guarantee us the same quality of oil continuously, I should like to know what it is. These tests convey a great deal of information as tests, but the next point is, how we are going to secure the same quality of oil.

Mr. GOODALE. I tried to cover that point in my paper by saying that I believe it is necessary in contracting for oil to have the contract contain a provision that each shipment of oil shall be up to a certain agreed standard; and I believe that each shipment received ought to be tested. The cost is very slight, and it would be of great value to the mill.

A MEMBER. How should we proceed to have our oil tested? When I tested oil fifteen or twenty years ago, I took a bottle of it to Brown University, Providence, but did not get a very satisfactory answer.

Mr. GOODALE. I don't know any better way than the way in which I made these tests. If I had known any better method, I should have adopted it. It is some little trouble, but it is well worth the amount it costs.

The PRESIDENT. If no one wishes to say anything more on this subject we will pass to the next topic, which is "The life of incandescent lamps," by Mr. W. S. SOUTHWORTH of Lowell, Mass.

THE LIFE OF INCANDESCENT LAMPS.

By Mr. W. S. SOUTHWORTH, Lowell, Mass.

In speaking on the topic assigned me, I take it that I am expected to state only the results of my own rather limited experience.

We have at the Massachusetts Cotton Mills one Edison dynamo, of a maximum capacity of about 600 twenty-candlepower lamps, though but 510 lamps have been installed. This plant has been in use about five years, and the dynamo runs whenever the machinery does, as the mill in which it is located is so darkened by the proximity of other buildings that a considerable number of lamps is in use even at mid-day in summer.

The lamps used are the Edison. As many other people have done, we began with sixteen-candle-power lamps, but later concluded the twenty-candle-power were more satisfactory for use in our weaving rooms, where we have one lamp to four thirtyinch looms. We have from time to time used small lots of lamps of other makes, experimentally, but have as yet found no sufficient reason for making a change. Until within a year we have ordered lamps of one or two vólts above the rating of the dynamo, as a result of which my figures as to their average life are no doubt a little more favorable than they might otherwise have been.

I find that the average number of yearly renewals is 276. For one year I had an account kept of the number of lamps burning at different hours of the day, on pleasant and on cloudy

days, in each month; and I assume that the average so ascertained holds good for the other four years. The total number of lamp-hours for that year is 431,220. As an offset to the fact that the lamps now in use should be taken as about half burned out, I add 51 to the number of yearly renewals (that is, one-fifth of half the whole number in use), making 367; using this sum as a divisor gives 1,319 hours average life of a lamp. Examples of very long life of individual lamps are common. For instance, we have one in the office safe which at this time has been burning something more than 6,200 hours, and is still in fair condition.

The largest item in the cost of operating the plant is of course for power. I estimate the average requirement at twenty horse-power, which I put at $25 per annum, making $500. Of course you will understand that water power is relied upon for a considerable portion of the year. For depreciation and repairs I allow $250; for interest, $225; lamps, $110; and for oil, attendance, etc., $75; making a total of $1,160 a year. The cost of gas saved, at $1.10 per thousand feet, I find to be about $1,500. This would make the cost of the electric light equivalent to that of gas at eighty-five cents a thousand. I consider that it has certain advantages over gas which have a money value, though that value is difficult to estimate.

The PRESIDENT. You have heard the paper of Mr. SOUTHWORTH. If the members would like to enter into any discussion of that subject, I should be pleased to have them. I suppose gas would come into the discussion as well as electric lights.

Mr. BROWN. My experience with electric lights has been limited. We are running only 100. Mr. SOUTHWORTH mentioned an incident which recalled an experience of my own, and that is, that he commenced with sixteen-candle-power lamps, but soon got to using twenty-candle-power lamps. I would like to ask what the depreciation of the lamp is. The average life of an electric lamp, as I understand, is 1,000 hours. Now when that lamp is burned one-tenth of that time, is onetenth of its light done away with? I think our lamps keep growing dimmer and dimmer, and I have sometimes taken them out before they were finished, because they were not giving light enough.

Mr. SOUTHWORTH. I think that is the common experience of everybody who uses incandescent lamps. Many people have come to me with lamps to be tested, claiming that they would not deteriorate or blacken. I have tried a good many experiments with them, and have not found that claim borne out. I should say that after a lamp had burned 500 hours not one-half of its value was gone; I don't think they deteriorate so rapidly as that. I rarely have any taken out; I think they generally burn until the lamp breaks, or until the lamp is smashed, perhaps intentionally, by the weaver.

A MEMBER. I should be pleased if some gentleman would tell us the comparative cost of the incandescent and the are light.

The PRESIDENT. Perhaps Mr. WOODBURY can give us some light on that point.

Mr. WOODBURY. I don't know as I can answer that, because the are light is so peculiar in its requirements. I suppose, for instance, that in the position in which they would be applied in an iron foundry incandescent lights would be almost worthless,