TWO public tests of the diffusion of light were given at the Walker Building of the Massachusetts Institute of Technology, Boston, on the 11th and 12th of September, 1900.

The representatives were present of all the different types of glass now available for this purpose which were tested, having been specially asked to make any suggestion or to take any exception to the methods adopted. None was made.

The conclusions which may be deduced from these tests, with a window twelve inches square serving as the sole source of light in a large hall with dark ceiling, are:—

First.—Windows of the customary height, but of one-third the width commonly adopted, when glazed with ribbed or suitable prismatic glass, will give on a bright day as much effective light as the full width of window glazed with plane glass; on a cloudy day or in a position where the light from the sky is derived from a limited area, even greater ratio.

Second.—Windows of the type common now in mills, workshops or school-rooms, now fitted with plane glass, if reglazed in the upper half only with ribbed or prismatic glass, will yield on a bright day more than fifty per cent excess of effective light, or on dark days a larger ratio. If reglazed down to, but not including, the lower panes (in which we advise plane-glass), the increase in effective light will be much greater.

Third.—Whether or not the increase of effective light will be as great in a room now fully lighted by the customary number of windows of plane-glass as in this hall lighted with a single twelve-inch window has not yet been determined, but it has been proved in mill practice that the light is much improved in quality and is rendered much more effective, both near to and far away from the windows.

In reviewing the photographs presented by Mr. Norton, a very great contrast will be observed between the very dark, even black, impression from the plane-glass and the very white impression of the diffused light from the ribbed or prismatic forms. This impression must be taken with some reservation, bearing in mind that but one window twelve inches square was used to light a large hall. If photographs were taken in a room already well lighted by numerous windows of plane-glass, then in the same room with the windows reglazed for diffusion, the contrast of white and black would be lessened, but the change in the impression would yet be a true measure of the increase in the effective light in the different parts of the room. We have as yet had no opportunity to take such photographs, but hope to be given one. If any of our members near Boston should decide to reglaze the upper sash in a room now fairly lighted with plane-glass, a notice is requested before beginning, in order that Mr. Norton may take photometric measurements of the light under present conditions and after the change.

At the meeting September 11th, the following remarks were made by me:—

Prismatic forms of glass and ribbed glass have been working a diffusion of light ever since they were made use of for glazing. The prismatic glass used has been mostly in fancy forms, but the ribbed glass has been used in the doors and windows of lavatories and other places, instead of transparent glass, to my own knowledge, more than forty years. Wherever such use has occurred light has been diffused or deflected. Prismatic forms of glass have been in use for the diffusion or deflection of light for more than fifty years in the decks of ships, especially over cabins. Here is one of the forms of prismatic glass, which has been used as a paper-weight by Mr. George Hayes of New York, builder of skylight, for twenty-eight years, having been previously in use over a ship's cabin.

It follows that, if any claims for letters-patent for the diffusion or deflection of light by the use of any type or form of prismatic glass have ever been granted, such claims or patents are now of no effect and may be wholly disregarded in a general dealing with this question. Special designs may be registered or patented as designs or forms capable of being applied to the deflection or diffusion of light, and also methods of framing or glazing; but these claims may be set aside as of no general effect or importance.

My own attention was called to this matter in the year 1883, when I inspected a large number of the modern English cotton-mills. I found them glazed with rough plate-glass of rather poor quality, the common glass of England being very inferior to our own, from lack of an abundant supply of sand of good quality. On asking the reason, I was told that the light within was more uniform and better. To my surprise this appeared to be true. The interior of the mill had the aspect of diffused light. It was also claimed that the extra thickness of the glass also served a useful purpose in maintaining more uniform conditions of temperature.

This fact led me to reason on the subject. I looked into the construction of the Fresnel lens, in which, as you are aware, a combination of lenses and curved surfaces concentrates the rays of light in a single, far-reaching beam. I reasoned that if by one set of angles or curves the rays of light could be concentrated, then by reversal, light could be diffused. I consulted the principal glass-makers and experts, suggesting the manufacture of panes of glass with one face

in prismatic shape laid out by computation for the deflection and diffusion of rays of light when set vertically in the windows of a factory. I learned that the first-cost of the plant for making such glass would be about five thousand dollars. I learned that the glass would be very costly, would require special methods of framing or sash and glazing; and that, under the conditions of a factory or workshop, it might be very difficult to keep the glass clean if the prisms were on the inside, and in a smoky or dusty factory-town it might be even more difficult to keep the prisms clean if they were placed on the outside. I concluded that if I were to attempt the introduction of such very costly methods for the general glazing of a factory, my own light as an adviser would be speedily extinguished.

I did not like to give up the idea, and I next gathered as many examples as I could find of window-glass of different types — fancy and prismatic glass, corrugated glass, curved glass, ribbed glass, rough plate and plane. Here are the types of glass which collected. I placed these examples in charge of Mr. Charles W. Hinman, then gas-inspector, instructing him to take photographic images of the light, which he did, placing a bright Argand gas-burner on one side of a slit about two inches by one-half inch. He submitted the photographs to me, which I now present to you.

This test disclosed the fact that for ordinary use in glazing factory, workshop or school-room, the best results would be attained by the use of the ribbed glass in true curves, inverse and obverse, twenty-one ribs to the inch, which has since been known as the "factory-ribbed," recommended by myself. Several of our members were induced to try this method of glazing, giving very favorable reports. We at first placed the ribs horizontally, but the line of bright light deflected towards the eyes of the workman may become injurious, and, although the diffusion from horizontal lines is a little better and more effective, the plan of setting is now to place the ribs vertically.

Having tested the theory in practice in a few cases, it now became expedient to bring the matter to a scientific basis and to prove to my scientific friends, who have treated my theory with absolute incredulity, that it was well grounded. They have treated my suggestions with polite discredit, having apparently assumed that I thought the quantity of light might be increased by the method suggested, when, in fact, my first hypothesis was limited to a better diffusion of the light transmitted. I therefore made the arrangements for the first tests made by Mr. Norton at the Massachusetts Institute of Technology. The last scientific observer who viewed the results a few days since, being thoroughly convinced by his own observation, exclaimed, "Why did not some donkey think of this before?" reminding me of another little invention of which a friend remarked that it was "so simple that nobody but a fool would ever have thought of it." These little accidental discoveries are sometimes of considerable importance, even when made by one who has no exact knowledge of the science which he applies.

Yet, this subject of diffusion of light by ribbed and prismatic forms of glass had been made a study, theoretic and practical, by two physicists of very high repute, resulting in the introduction of the Holophane glass shade, now represented by the Holophane Glass Company, No. 1 Broadway, New York. The effect of this shade, when put upon a Welsbach or powerful electric light, is as startling and as effective as the use of the same types of glass in windows. I have suggested to the Holophane Glass Company to make a bowl-shaped shade that may rest upon the font of a Bradley & Hubbard standard kerosene oil-lamp, deflecting light upon the book or paper, casting a very full light upon the ceiling and thence reflecting it, while at the same time throwing the heat away from the head of the reader toward the upper part of the room.

This theory being tested empirically in a few cases between 1884 and 1892 led to the organization of several different companies for promoting the diffusion of light, each of which has of late been urging the adoption of its method, with a view to profit. These undertakings, in competition with the ribbed glass rolled in large sheets, first recommended by the undersigned, began with prismatic glass of very perfect quality, but of necessity cast in small plates, involving heavy expense in framing and glazing. Under the gradual competition which has arisen, rolled glass in large sheets of prismatic form with sharp angles, corrugated glass and other types are now offered at very moderate cost in competition with the improved ribbed glass now in the market. In view of the large number of competitors and the claims made by the representatives of each, it became expedient to make this final test of their relative merits, in order that the members of the Associated Factory Mutual Companies might be put in possession of all the information that is in the power of their agents, the officers of the several companies, to give them. The representatives of the various types of glass now about to be tested are present and are urgently requested to make any suggestions by which the tests may be made more complete and impartial.

It will be made plain in Mr. Norton's report that, while the adoption of the factory ribbed glass will serve the general purpose of diffusion as well as, if not better than, any other type, especially in view of its low relative cost, yet it is now evident that there are many places where the prismatic forms will be more effective, either in the rolled plate or in other instances, as in canopies, in the finest types of cast glass.

In closing windows against fire, either the "Maze" or the ribbed wire glass may be adopted, assuring safety and at the same time improving the light within.

Care should be exercised in setting thick wired glass in metal frames. The lower edge must, of course, bear directly on the frame. The sides and top should be fitted loosely, so that the differential expansion and contraction of glass and frame may not crack the glass.

At present no general rules can be established, but if any of our members desire expert assistance, the services of Mr. Norton may be availed of at moderate charge, including expenses, so far as such work may be consistent with his duties at the Massachusetts Institute of Technology. Applications for such special service may be sent to any of the Factory Mutual Companies or to the Bureau of Inspections. After a little study our special inspectors will be prepared to advise on this matter. In many cases where there is any doubt, a plan and elevation, showing the relation of adjacent buildings, the width of alleys, etc., with points of compass, may enable us to advise what type of glass to use, without the necessity of a special visit. Applications of this kind will be referred to Mr. Norton, and will be subject to a reasonable charge on his part for such consultation.

It seems rather singular that there is no record, as far as I know, of any previous effort to take advantage of this principle of diffusion of light in glazing vertical windows and improving the effective light of the factory, the workshop or the school-room. A few years ago an effort was made to introduce prismatic glass, in order to gain improvement in the lighting of city shops and offices. This undertaking originated in Chicago, with parties whose names are unknown to me. On coming into competition with the ribbed glass, these parties attempted to throw ridicule upon our methods and to discredit Mr. Norton and myself, even by writing very impertinent letters to the officers of the Massachusetts Institute of Technology, where these tests were made. I am informed that the parties then undertaking this work have either failed or have ceased to be a factor in the case.

It will be observed that, although these tests have always been made in these buildings, as the present tests are now being made, neither the corporation nor the professors of the Institute of Technology have anything to do with the matter. The use of this room and other rooms has been granted for the purpose of making the tests, and Mr. Norton is wholly in the service of the Factory Mutual Companies, represented by myself.

This opportunity has been offered to the makers of every type of glass of which we can get any knowledge. Some have contributed not only the glass, but towards the expense of these tests. Others have contributed the glass; others have omitted to take any part, but examples of their glass have been obtained and will be presented. We shall not indicate to you in any way any distinction among the representatives whose glass will be tested; each type will speak for itself. The records will be made by photograph and photometer, and when our final report is made we hope to make it exhaustive; yet there may be one type of glass to be made as the result of what we have discovered.

It may be judicious to glaze each horizontal set of window-panes (from one to six rows in the ordinary factory window) with glass that will deflect all the rays that fall on each horizontal line to a different section of the ceiling. This can be done, as will presently appear, to a distance of over 60 feet from the window. We may thus throw all the rays of light upon the ceiling, securing diffusion by reflection, doing away with all window-shades, and also, perhaps, removing the objection which obtains to the deflection of light on the line of the eyes of the operative. We may be able to place the lines horizontally and yet not bring any glare of light on the level of the eye. One of the advantages claimed in the English mills, in the use of the rough plate-glass, was the removal of the need of window-shades, which are somewhat hazardous in case of fire, spreading fire rapidly. Our glass is so clear and our atmosphere so much more translucent, that this hope has not been justified by the use of ribbed glass in this country.

We feel very sure that certain types of corrugated or prismatic glass, if not already made, may be made for the special purpose of glazing or reglazing the upper sash in school-rooms of 28 feet from window to wall, by which the rays of light may be deflected to the ceilings, which may be kept white. The light may then be reflected from the ceiling to the desks, without any danger of any strong lines of light injurious to the eyes being thrown upon them. If this can be accomplished, no cloth shades will be needed, even on the south side.

In the study of the science of diffusion of light it has become apparent that there are two phases of the question, each of which must be considered according to the conditions.

In the isolated factory constituting the majority of the works dealt with in the Factory Mutual System, the area around the factory virtually gives a hemispherical source for the diffusion of light, and in such cases that type of glass which works a sufficiently effective diffusion at the least cost is the one to be selected. On the other hand, in dealing with windows on narrow streets, alleyways or in basements, as in crowded streets, the majority of instances call for special treatment, with a view to deflection as well as diffusion. Another type at higher cost may be rightly adopted. Hence the need of special consideration being given to each aspect of the building, according to its position and the relation of the windows to large or limited sky area.

It may be judicious in all cases to glaze the lower row of panes in factory, workshop and school-house windows with plane-glass, so that the unpleasant impression of not being able to look out may not be given. It has sometimes been suggested that some portion of each window should be of plane-glass for hygienic reasons, upon the ground that "sunlight transmitted through plane-glass has a greater hygienic value than that which is refracted through a denser medium."

The roughened glass of various types is not a "denser" medium. The glass only diffuses or deflects the rays of light. The hygienic value of this diffusion in its effect upon the eye, when the plates are rightly adjusted, cannot be questioned. It has also been observed that the heat of the sun is diffused, as well as the light, making the warming of the room more uniform. There is yet much work to be done on this line of observation.

Measures are being taken to establish permanent apparatus for testing plates and showing the diffusion of light in the laboratory of the Bureau of Inspections, 31 Milk Street, Boston, where any member may observe the results.

This apparatus will be very simple and of small cost. If any members or experimenters desire to set one up for their own use, the diagrams will be supplied. It may happen that such an apparatus should be set up in print-works, color mills or dye-works, for testing glass in its influence on colored work, as well as in diffusion, as it is necessary to use the whitest glass in such works, the dark greenish tinge vitiating the colors in the work.

I hope these facts will secure attention from School Committees and Superintendents. The light in school-rooms should be wholly on the left of the pupils, and the present customary area, 32 feet in length and 28 feet in width, is a very bad form for lighting effectually on the inner part of the room from plane-glass. It also occurs to me that the effect of diffusion of light, probably of heat also, upon the growth of plants, perhaps preventing scorching, might well be tested by the owners or occupants of greenhouses.

Proofs of the text of this report have been submitted to the representatives of each type of glass tested or named, with the request that any exceptions should be taken or suggestions made, to which suitable attention would be given. No exceptions have been taken, but some verbal changes, making the text more clear, have been adopted. Respectfully submitted, EDWARD ATKINSON.

BOSTON, MASS., October 8, 1900.