rubbing friction of its surface against the interior of the flue, heat, which produced a sufficient amount of steam for the dyeing, heating, and boiling purposes of the establishment, so that the heat for the whole works was generated by the water wheel.

Mr. Jerome Wheelock.-I think if we continue the discussion of compound engines far enough we shall make out that you can saw wood with a claw-hammer (laughter). I am going to say in passing that in several instances I have been called upon to furnish engines, condensing at one end and non-condensing at the other. I think in one or two instances the parties were so disgusted with the performance that they abandoned the exhaust heating and took the steam from the boilers for the heating with very marked success. In every instance where steam is used and where water is plentiful, I think the condenser can be used with great advantage, and use live steam for heating, and with a properly constructed engine of proper size where the cut-off can take place early, for ordinary purposes you can get results which will compare favorably with a compound engine. The compound principle is an apology for leaky valves and pistons. That is antagonistic to the general motion, but I put it in.

Mr. T. J. Borden. My experience in manufacturing convinces me that power can be derived from water under favorable conditions, with greater economy than from steam under the best of circumstances, except in the immediate vicinity of coal mines, provided the quantity of steam required in the processes of manufacturing is not large in proportion to that needed for power.

In a few branches of manufacturing, the steam required for power is not more than one-half or one-third of that for other purposes. In such cases, power can be produced with a steam engine, the exhaust steam of which may be used for the other processes almost as effectively as if it had not passed through the engine. Under such conditions water is of little or no value for power. In by far the greater part of manufacturing operations, the steam required for power is large, compared to that requisite for other

purposes..

The important elements of value in water powers are:

1. Water supply that can be utilized by a moderate outlay for dam and water ways in proportion to the power the stream will yield.

2. Facilities for holding back considerable reserves of water in lakes for use in the summer and early fall months.

3. Location easily accessible to the markets.

Water powers possessing these characteristics, and there are many such in this country, are of very considerable value as sources of power notwithstanding the great reduction in recent years in the cost of producing power by steam. Streams possessing no reserves of water under control for use in dry seasons cannot be utilized to so large a proportion of their average flow as those having such reserves.

The power to be relied upon from streams of this character is that which can be developed by about a minimum flow of the stream, with a fall equal to the difference in height between the surface of water above the dam, and at the tail race in times of ordinary freshets. The leading source of disappointment in the use of water powers, is the placing of more machinery on a given water privilege than the stream is capable of driving in times of a minimum flow of water.

This is no good reason for condemning or underrating the value of water powers.

Mr. H. H. Suplee.-In connection with Mr. Durfee's remarks about generating heat from water power, I think it might be well to remember that the Cowles Electric Smelting Company at Lockport are producing exceedingly high temperatures with dynamos driven entirely by water power, and that those high temperatures, which are probably the highest that are used in the arts, are obtained without the combustion of any fuel whatever. They are using electric furnaces for the purpose of producing metallic aluminium from corundum.

Mr. Borden.-I might say in connection with the water power at Holyoke that the original owners of the water power derive a decided profit from it; and that the reason why it is not now sought for is because the power is exhausted during a consider. able part of the year. The party who now undertakes to do anything with water power in Holyoke must do it, taking the chances of running short in the summer season, and that is the principal reason why steam is used instead of water power; not because it is cheaper, if the water power could be had through the year.

Prof. Thurston.-I have not arisen to take serious part in this discussion. Before leaving home I had not time to read this paper with the care I wanted to give it, and I may ask the privilege of putting in my remarks in writing, if I can get time between the adjournment of the meeting and the publication of the paper.

I have had some experience of late in the use of water power, and in the comparison of water with steam, and I have been impressed as I never was before, with the very serious unreliability of water power in many sections of our country, both as to the amount flowing in the course of the year and securing the power that it may give, at any one time, economically. My observation. has been, where I have had the opportunity of observing the flow of streams during the last fifteen or twenty years, that, at every point in the country with which I have been familiar, the flow of streams is becoming rather less annually, and vastly more irregular continually, and recently we have been endeavoring to harness a stream which has proved to be the most unmanageable case of the kind I have met yet. It simply emphasizes the fact, which I think is observed everywhere, that one of the reasons for substi tuting steam for water is that we cannot rely on having water power when we want it; and, although the stream may give full power for nine months in the year, for three months it is apt to be of very uncertain flow as well as of very small volume. The stream of which I now speak has a flow, I should imagine, at certain periods in the year, of not less than a thousand horse power. At other seasons, I should presume that the amount of flow would hardly be enough to supply the ordinary leakage from flume and wheel. In this case, the objection to utilization of water is not the cost of fitting up. I can, I think, under these specially favorable conditions get the dam, and the flume and water-wheel, the whole thing in running order, at a cost of not over $40 per horse power. I do not think that it would cost that; that is to put the whole thing in running order. When that is done we have to contend with this uncertainty as to flow. We may have a dry spring and a wet summer, or a dry summer and a wet fall, or a wet summer and a dry fall. We never know what to expect from this stream, except that through the winter we can get all the power we want.

Another difficulty that I have found in the application of water power, and one which has led to the changing of plans in the direction of substituting steam, is the difficulty of securing exact regulation of water power. I have not yet found a governor that could handle our water power satisfactorily, that would keep the speed anywhere near where it ought to be in driving electric light apparatus. I doubt very much whether, in the large majority of eases, it is possible to secure that regulation necessary for doing

that kind of work. Where the work is constant, as in a cotton mill, there is, of course, very little difficulty in handling the water power satisfactorily in this respect; but where large amounts of power are liable to be thrown on and off, as in a system of electric lighting, I find no governor of any service.

It is partly because our streams are getting less and less reliable all the time that we see this enormous increase of steam power throughout the country. There is, of course, an enormous decrease going on continually in the cost of steam. It has become financially practicable to-day to put steam in a mill at Holyoke or Lowell, and to neglect the water power which lies right at hand. I presume the time has come, in many sections of the country, when the cost of installation is so great with water power and so little, comparatively, with steam, that the difference in the actual cost of power becomes so slight, that the special advantages of steam will more than compensate for those differences in first cost, and that steam may be introduced where water power can be had for a comparatively small cost. The cost of steam, when used in the manner suggested in the course of the discussion, and the exhaust employed for other purposes, is enormously reduced. I do not agree with my friend Wheelock, if he means to put his statement as a general proposition, at all. I should presume that, in all cases that are not especially unfavorable, it would be very wise, if possible, to save that exhaust steam and use it. I have never known a case in a well arranged plan where that was not the fact. I think that where we can get a horse power for three-quarters of a pound of steam per hour, as competing with power at four to eight times that amount, there can be very little question as to the wisdom of making use of exhaust steam for heating.

Prof. Denton.-I want to ask a question that has always puzzled me. An instructor desires to answer the question of water against steam very often, and certainly the written account of water versus steam is very mixed. I think in the Vienna report there was a statement that water power was infinitely cheaper than steam, giving a certain quotation from a Philadelphia engineer, and it was immediately answered by that same engineer stating that the quotation was entirely erroneous, because he had not taken account of the cost of repair to dam and wheels. When he put in those figures the cost was equal between the water and steam. That is the first instance I remember. Then the neces

sity comes in of having steam in the cotton mills for operations. other than to supply power, and this question arises, which I hope Mr. Borden will answer. Is it a fact that the mills using steam entirely can compete in the market with the mills using water? If they do, why is it that the less cost of the water, which I gather from his remarks is a fact, does not enable the mills using the water to put the market down against the mills which exclusively depend on steam?

Prof. De Volson Wood.-In regard to water power and steam power, the uncertainty of the water power is a sufficient reason for bringing steam into use, and the statement which was made by the last speaker is sufficient to settle the point.

Prof. Denton.-Not where there is plenty of water and no drought.

Prof. Wood. Then the question was not just as I supposed. But I was going to make an illustration which would be partly to the point in the discussion. It is cheaper to transport freight on a canal per ton-mile than it is on a railroad; but the railroads are drying up the canals. Why? Because time is a great element, and if the water power flows only a part of the year, and fails when it would be most profitable to turn out its products, the use of a steam engine, if at hand, might decide the question as to whether that manufacturer relied upon steam or water; for if he relied upon the steam engine he might continue its use and let the water run to waste. In regard to the figures in the table on the sixty-first page:-In attempting to determine the efficiency of plant, I find an omission which makes it practically impossible for us to determine the efficiency of plant, and that is the heating power of the coal. Now for the purpose of this paper, perhaps it may be unnecessary to more than state the amount of coal used per indicated horse power per hour; but there is such a difference in the heating capacity of coal-whether it be coal containing a great deal of dust and dirt, or whether it be of a good quality, or still further, if it be picked coal, that if we attempt to determine the efficiency of plant it becomes necessary for those who report these figures to determine, either chemically or otherwise, the heating power of the coal. Not long since, I think it is within a year or so, a triple expansion engine on a small steamer was put afloat in England, and the makers were required to develop an indicated horse power for one and a quarter pounds of coal, and as a result of the trial it was reported that they used 1.23 lbs.,