Page images
PDF
EPUB

reduction of gas in candle feet. There is a wide diversity of opinion with regard to the amount of oil used per thousand feet to produce a given candle power of gas. I am very much in favor of hourly or half-hourly observation of the photometer in connection with the amount of gas made. If it can be demonstrated that there is 4,200 candle feet in a gallon of oil of certain quality, I think that the measure of the production of gas in candle feet will tell us what we are doing with our oil-whether we are making gas with it or making lampblack of it.

On motion of Mr. Boardman, a vote of thanks was tendered Mr. Clark for his paper.

FIRST DAY-AFTERNOON SESSION.

Mr. A. C. Humphreys, M.E., Philadelphia, Pa., read the following paper, entitled :

ILLUMINATION VS. CANDLE POWER.

I have undertaken to prepare for this meeting a paper under the title of "Illumination vs. Candle Power." The subject of measurement of light is a large one, and I cannot hope to more than touch upon certain special features of the subject. The title will suggest that I have in mind some of the papers recently written in this line, namely: The paper by Mr. Boardman,* at our last year's meeting; the papers by Messrs. Prichardt and Taber, at the last meeting of the New England Association; and the paper by Mr. Chollar,§ at the last meeting of the Western Association.

Mr. Boardman told us, you will remember, of the increase from sixteen to twenty-two candle power he obtained by substituting a student's Argand chimney, for a regular Argand chimney, but that while he obtained this apparent advantage the "illumination" was actually less, as shown by applying the chimney to an Argand at his house. In the paper and the following discussion you were told that a white light would not "diffuse" as well as a yellow light. It was also stated during the discussion, and with great confidence, that a sixteen candle coal gas gave as much light, on account of this "extra diffusibility," as a twenty-candle water gas-the twenty-candle

* Vol. VII., p. 301 of Proceedings. + AMERICAN Gas Light JourNAL, Vol. XLVI., March 2, pp. 130, 140. § Vol. XLVII., July 2, '87, p. 7.

[ocr errors]

water gas being whiter than the coal gas; and in explanation of this it was suggested that the whiter light from water gas was intense, local, but not diffusive like coal gas."

In Mr. Prichard's paper considerable stress is laid upon a law of the intensity of light to the effect that "rays emitted obliquely from a surface are less intense in proportion as they were more inclined to the surface which emits them," and then, that by this law, "the rays of light from a flame as they leave the horizontal and approach the vertical, leave the maximum measured candle power, and approach the minimum of zero." We were also told that flames must be of the same size and same color to enable us to measure their lighting power correctly. The old question of the opal globe was also brought up prominently. It was also proposed that daylight should be regarded as the standard—that is, a white light. Also that a sixteen-candle gas, which was whiter than the richer twenty-candle, on account of its increased "temperature per unit of flame areas," produced a better light than the richer gas-of Boston. A disagreement from what has gone before is here to be noted.

By Mr. Taber's paper we were reminded of the uncertainty of our photometrical measurements. The varying quality of the socalled standard candle, the necessity of bars of varying length according to the strength of the light, the lack of accurate formula for correction of candle and gas consumption, the question as to how to read the disc, the color trouble, the abnormal condition of atmosphere-fog, for instance-entering as a disturbing element. And also, the opinion is expressed that the subject is one for the physicist, instead of the gas engineer. Then, by a carefully prepared table, it is to be noticed that the candle power does not follow the amount of bromine illuminants, and that CO is a positive evil. In the discussion which followed, the example of the illumination from a bull's-eye lantern is used to demonstrate the varying power of the rays emitted by a gas flame, according to the angle they make with the surface of the flame. Also, we were told that candle-power "means nothing." Mr. Taber found that Mr. Boardman's experiment gave him an increase in candle-power of from 16 to 19.44. Mr. Prichard had found

no difference whatever.

Mr. Thomas spoke of the relative values of oil lights and gas

lights-compared directly, I presume-and that the "illuminating duty" of one eighteen-candle gas flame was found to be far in excess of two twelve-candle oil lamps. This was the result of a careful test-presumably with a bar photometer. The experience of Dr. Morton in measuring so-called 2,000-candle electric lights is referred to, and we are reminded that he only found them to be from 500 to 1,100 candles; and the opinion is expressed that if a photometer had been used in these experiments, the light would have largely exceeded 1,100 candles. Also that an observer looking at a thirty-six-candle and a nineteen-candle gas as ordinarily burned, would have difficulty in distinguishing between them; and, still further, the belief was expressed that a nineteen-candle gas would afford a better illumination than a thirty-six-candle gas. Parenthetically, I would here like to say that if the gentlemen would guarantee to convince our consumers on this point, I should be only too glad to accept the theory, anyway during business hours, and let the statement go unchallenged.

At the close of this discussion Messrs. Prichard and Taber were appointed as a committee to continue their investigations and experiments during the following year.

In Mr. Chollar's paper we are told to beware how we put our faith in the law which says that the light from a luminous body decreases in proportion to the square of the distance—that is, the law on which photometry is based; and he also refers to Professor Tait's statement of the law, namely: "If the medium be transparent, the intensity of illumination which a luminous point can produce on a white surface directly exposed to it is inversely as the square of the distance."

Attention is then drawn to the fact that the well-known law is simply following out the principle that the surfaces of spheres are to each other as the squares of their diameters. Then, the proof is attempted of the proposition that, "irrespective of intensity, the quantity of light from a luminous body at any particular point is independent of the distance of the body, and is in direct proportion to its diameter;" and it is stated "The rule governing the value of lights, therefore, would be something like this: The light from a luminous body is inversely as the square of the distance, and directly as its projected area." We are also

advised in the study of the subject to drop empiricism and go at the matter as thoroughly and systematically as the electricians do in their measurements; and we are encouraged to believe that we shall be able, by means of high candle power lamps, to compete with the 1,200 and 2,000-candle power electric lights.

These papers you have doubtless all read, and it may seem that it was unnecessary for me at this time to refer to them at such length. My object has been to bring clearly before you the fact that there is in the minds of the gas engineers of America much uncertainty on the subject of light and its measurement, and to demonstrate the necessity of a special consideration of this subject on the part of this Association. I take it for granted that we are all gas engineers-not coal gas engineers, or water gas engineers, or oil gas engineers, or wood gas engineers, or any other special kind of gas engineers. Perhaps it would be better if we said we were light engineers. If we cannot go this far, it is high time we went far enough to be able to say that in this Association we are gas engineers, and that we are ready to study any subject in connection with our business, so as, as far as possible, to get at the exact truth, let it strike where it will. I believe most of the members will acknowledge there is room for improvement in this direction, though I am, for one, glad to acknowledge there is every appearance of a movement toward the overthrow of prejudice. I cannot expect to try to meet all the points raised in the papers referred to, and I know you would be sorry to have me attempt it. I believe, however, that I have been able to make certain experiments which will tend to clear up some of the questions raised.

It certainly is very deplorable if we must acknowledge to the public that we can tell actually nothing as to the value of the light we offer to them; and that is what we must practically acknowledge if we leave the subject at this point. Of course, the foundation laws of intensity of light which have governed us in our photometric work are:

I. The intensity of illumination on a given surface is inversely as the square of the distance from the source of light.

II. The intensity of illumination which is received obliquely is proportional to the cosine of the angle which the luminous. rays make with the normal to the illuminated surface.

It is with the first law we are chiefly concerned, for we can provide against the second law interfering in our photometric work by having the disc at right angles to the rays from the light to be measured; and this can be practically done, even if two candles are used, by the use of a bar of sufficient length.

I think that Mr. Prichard, where in his paper he refers to the rays emitted obliquely, has this law in mind, and most of his troubles appear to come from a misconception of this law. The law has nothing to do with the emission of light from an oblique luminous surface, but refers to the illumination of a surface oblique to the rays, and simply shows that mathematically, by reason of the incline of the surface, a greater area is exposed to the rays. Therefore we spread a certain number of rays over a greater area than in the case of a perpendicular surface, and consequently the intensity of illumination is diminished. And the exact measure of this inclined surface, as compared with the perpendicular surface, is found to be proportional to the cosine of the angle made by the ray with the normal to the illumined surface. This, therefore, need not come in to bother us. We only have to take account of it so far as to see to it that our photometer is adjusted to reduce the error from this source practically to zero. Part of the confusion on this point may have come from the fact that a flame is not entirely transparent to its own light, as shown by the difference in readings shown between the flame on edge and the flame on the flat. On this point there has been great diversity of opinion, and the rule, I believe, somewhat generally followed has been to take the reading from the flame at 45° as giving the average.

In my past experiments I very quickly satisfied myself of the incorrectness of this as a general rule, and, therefore, in connection with this work I assigned to one of my assistants a somewhat elaborate set of experiments to determine what angle gave the mean reading, and this with different qualities of gas. The following table, prepared by the assistant referred to (Mr. C. Russell Collins, M. E.), shows the average of the mean readings, or the mean angle of mean light to be 4.68°, and that this mean angle varies with the candle power of the gas. The differences of candle power were obtained by mixing different percentages of fuel (non-illuminating) water gas and illuminating water gas.

« PreviousContinue »