coal used? What kind? At what price? I have submitted herewith a table of questions which may lead to the establishment of such a unit.

When I came home, I will say to my English friends here, I was very glad to find that American inventors were working in the same direction. The Roney system, coming east from the far west, taken up by the firm of Westinghouse, Church, Kerr & Co., is being established in this neighborhood. It is another method of accomplishing the same purpose. If you will go to Curtis Davis & Co.'s works in Cambridge you may see a plant not yet completed, where they are consuming our ordinary Pocahontas coal without any smoke from the chimney stack, which is twenty-five feet in height above the roof of the boiler house. It is of iron, and upon the outside of it I am told you can bear your hand. It is a mere prima facie case I put before you as far as this matter is concerned. My intention is to keep informed in regard to it, and I make these suggestions for you to follow out in detail. By this American invention there is a complete combustion of fuel; there is claimed to be a large evaporation of water, and it is claimed to be a method that will rival the English method. Between the two you have an accomplished fact. Of course all the details must be gone into by experts who know the subject, as to the evaporation of water, consumption of fuel, repairs and everything else. The investigation I am making is merely preliminary. The subject is urgent in Great Britain, very ripe here, for Boston is beginning to be infested with coal smoke so badly that the city council, while I was away, passed an ordinance forbidding the use of smoky fuel without some kind of smoke prevention or smoke consumption.

In the matter of smoke consumption, that subject was treated at the British Association for the Advancement of Science, at which the Sennett system was presented, of which I expect full diagrams and description. It works the consumption of smok made by imperfect combustion and is apparently very succes ful. It is more likely to be successful in a locomotive engi

and in steamers than in plants like ours. If you can prevent smoke by complete combustion you will have no smoke to be consumed. Again, when I got home I found an American invention which corresponds very closely to the Sennett system. This apparatus is made by the Standard Smoke Consumer Company, 54 Lincoln Street, of which my friend and fellow townsman, Mr. Theodore Pinkham, is president. There is room for both, because there are many plants where perhaps smoke may be consumed with greater economy than it can be prevented; but in all the large plants the line of economy is in the direction of smoke prevention rather than smoke consumption. Smoke prevention is an accomplished fact in England, and is working with economy in labor and economy in fuel.

The common boiler used abroad in mill plants is either of the Lancashire or Galloway type. At first I had the impression that the Vicars apparatus might not be applicable to other types of boilers, and therefore I sent on to the Edgemoor Steel Works, in which they are manufacturing the Galloway boiler with great success, and obtained from them a complete set of drawings. Their type is different from the English. These drawings will be at the control of the government. But I do not think it will be necessary to limit the application of either of these methods of smoke prevention to any particular type of boiler. I have reason to believe that either the Vicars or Roney system can be applied to the kinds of boilers you now have as easily as to the Lancashire.

However, when you come to enclose your boilers in a nonheating medium, with a moderate draft, you will be free from some of the objections of contraction and expansion, and you may return to the types of boilers in which there are the fewest joints, the fewest tubes, and which are the least liable to leakage; that is, the old types which preceded the tubular upright boiler.*

Since this address was given, the attention of Messrs. Wm. Sellers & Co. of Philadelphia has been called to the Vicars stoker, and I am informed that they have given an order for the apparatus for trial, with a view to undertaking its introduction in this country if it proves to be a success.

It was stated that I should say something about the transmission of power, but I shall not go into that subject now. I found the Messrs. Coats were using rope transmission, and that is the practice in all the larger and newer mills of Yorkshire and Lancashire. Our friend JOHNSTON says it is perhaps because they don't know how to make a good leather belt. I shall not go into that. I was inclined to believe in rope driving five years ago; I am inclined to believe in it more now than then, but I have no title to pass judgment upon it. I think the Englishmen would have procured good belts if they had not found a preference for ropes. The most scientific man in the world that I know of connected with cotton mills, Dr. Bowman, is enthusiastic for rope transmission.

I am told that there is a book describing the Curtis Davis plant, which can be had at the close of the meeting. I am authorized to invite you to look at their works at Cambridge. If you go there, I want you to look at another matter. Two of my fads which are bound to come into general practice within a short time are there applied. One is their new building, built upon the plan that the mill shall hold up the wall, not that the wall shall hold up the mill. You can put up a timber frame holding itself together, and you may put off your decision as to what the outer covering shall be until you have placed the shafting and machinery if you choose; you may then put on your outside wall to serve as a window frame. The building at Cambridge to which I call your attention is the glycerine house. That glycerine building is planned exactly on the lines of my circular No. 10, issued to you about ten years ago.

ission
get

You will find at these works also, on the south side of the building, rough plate glass. Before I went abroad I left instructions for experiments for determining the tra and diffusion of light. I thought it might be necess some new types of ribbed glass, but I am satisfie kind, of which here are several samples, is well

that no other kind is necessary. You will be surprised to know that the amount of light that is absorbed by this ribbed glass is ten per cent., and by the clear glass is eight per cent. There is only two per cent. difference in the obstruction of light between these samples which I have here. One is a double-thick German clear, the other is a fine-ribbed plate of same thickness. Now, through that plain glass you get no diffusion; your light falls directly on the floor, and the clear sunlight falls upon your looms or frames, making dark shadows; while with the ribbed glasses you get a diffusion through the room without a glare, you need no window shades and you have no dark shadow. Now, then, in the practical use of light, you will get a vast deal more with that ribbed glass without window shades than you will with the clear glass with the window shades which you must have on the south side. Moreover, it will pay you to put in thick glass rather than thin, in order to save the heat in winter and keep out the heat in summer. But I have only begun on the alphabet of this matter. The hammered glass is practically the same thing, but is not as good as the ribbed glass in the transmission of light in horizontal planes. I have a table here prepared by Mr. Hinman, giving the relative loss of light, which with ribbed glass is ten per cent., and with clear glass, double-thick, eight per cent., single-thick, seven per cent. The dispersion of light through the ribbed glass is relatively sixty-four degrees, as compared to 0 degrees in the plain glass. The relative cost per fifty feet is for one-eighth-inch ribbed glass five dollars; double-thick German, three dollars and a quarter. On the south side of Curtis Davis's laboratory you will find rough plate glass. I think there is room for a further development of the science of glazing, which is of great importance. I tried to find out if there was any science of glazing in England, where they use rough plate glass which is not so good as ours. They have no glass sand as good as the Pittsburg or Berkshire for the diffusion of light. At the meeting of the British Association an explanation was given of the Fresnel lens for trans

mitting light in one beam horizontally. Now, by reversion glass apparatus has been made to serve as a diffuser of light for an arc lamp. The inventor worked on my theory of window glass. He has applied this idea to a conical glass which is put under the arc light; with this he has done away with the glare and destroyed the shadow. You may have daylight from an arc light in your mills without the shadow and without the glare. I have ordered one of those lamps to be sent to me.*

* The following table will give the approximate results of a preliminary investigation of various types of window glass procured from Messrs. Hills, Turner & Co, which has been conducted by Mr. Charles W. Hinman since I left home. It is proposed to extend these investigations at a later day by the application of photography, which will give a more exact measure of the relative transmission of light than the methods formerly used.

What is called the ribbed glass has about twenty-two very fine ribs or flutes to the inch; No. 1 fluted has four deeper lines or flutes to the inch; No. 2 fluted eleven lines or flutes to the inch. The rough plate glass you are doubtless familiar with. Each of these types of glass has been justified in use in weaving rooms, laboratories, engraving rooms and the like. The point at issue is, which is the best?

The glass chiefly used in Great Britain is the rough plate. I think it is not the best. Neither, as I have stated, does their glass approach ours in clearness and translucency, because they have no sand equal to our own which can be applied to common purposes. You will observe that the conductivity of heat is in the ratio of forty-seven in the use of the ribbed glass one-quarter inch thick, as compared to fifty-six with the use of the ribbed glass one-eighth inch thick, or the double-thick German, while the obstruction of light does not vary very greatly; it is ten per cent. in respect to the one-eighth inch, nine per cent. in respect to the one-fourth inch, which appears to be a more translucent quality of glass than the other eight per cent. double thick.

The relative price per fifty feet: seven dollars for the one-quarter inch, five dollars one-eighth inch, three dollars and twenty-three cents double-thick German. Whether or not the saving of heat would pay for the difference in price I am unable to compute. I am inclined to think it would.

With respect to the diffusion or dispersion of light there can be no question. The ribbed glass takes precedence. If it is judicious to glaze a mill with glass as heavy as one-quarter inch, then manifestly the preference must be given to the solid window frame as distinguished from the ordinary moulding and boxed sash, the upper light opening inward on a transom. I have made no computation of the relative cost of window frames, but one would infer that the solid fixed frame with transom window would on the whole be better and perhaps cheaper than the common frame with the boxing, pulley stiles, weights, etc.

As this matter of glazing is one of paramount importance, I will venture to add a word, while correcting the proofs, to what I said at the meeting.

In connection with my son, who is an architect, Mr. William Atkinson, I have been planning a school-house which shall be as well lighted, as uniformly and satisfact heated and as thoroughly ventilated as our best factories now are. One ca reason why the requirements which must be met in the cotton factory in