criticising this off-hand. Your quotations from the English authorities and what they recommend are undoubtedly correct, so far as they go; but there is not a battery in England so big as ours, and the authorities you name have not had any experience in commercial battery work.

MR. PERRY: In England, the station load factor is from fourteen to eighteen; in this country, it sometimes reaches forty. A storage battery might be economically operated under a low line, as England, but not here.

MR. EDGAR : In England, they have no motor business; we have it. I am not speaking of your curve as impossible for England; but it is impossible for existing business in this country to-day.

MR. PERRY: I have assumed a curve that would be peculiarly favorable to battery storage. With this

curve, partial storage would undoubtedly be economical; but, even under these favorable conditions, complete storage would be out of the question. I have assumed certain conditions, and have tried to meet them by different methods, and have compared these methods by their relative cost. The curve I have assumed would be an unusual one for this country at the present time. For England,

the load line would be a good one.

THE PRESIDENT: This discussion is important and interesting, but we shall have to defer its further consideration until another time.

THE PRESIDENT : The next paper on our programme is by Messrs. Houston and Kennelly, on "A New Method of Measuring Illumination," and will be read by Professor Houston.

OF MEASURING

A NEW METHOD OF

ILLUMINATION.

In almost all cases, the object of artificial lighting is to provide a sufficient intensity of illumination upon the illumined surfaces. In practice, the illumination is generally required in order to permit some kind of work to be carried on. In any case, the character and intensity of the illumination will necessarily vary with the nature of the work. In a room in which reading only is to be done, it suffices if the proper illumination is obtained on the pages of the books, that is, on the reading tables, though the remainder of the room is comparatively dark. Where machinery is to be operated, the main light is generally required upon certain moving parts. In cases, however, where the character of the work requires the attendant to move about, or to observe some distant part of the machinery, a necessity exists for avoiding marked contrasts in the illumination of the room. Here, it is clear that the use of a single source, such as an arc lamp, is undesirable; since, apart from the fact that it would produce a powerful illumination in its immediate. neighborhood, it would necessarily cause annoying

shadows which could be avoided only by diffusing and scattering the light. It is evident, therefore, that in practice it is the illumination, rather than the luminous intensity of the sources, that is desired to be obtained and measured.

Although experience will enable the electrical engineer to meet the probable number and grouping of sources of light, such as incandescent lamps, that will be required to produce any desired illumination in any given space, yet it is well known that no rigid rule can be laid down as a safe guide in such cases; since the character and color of the surfaces of the walls and ceiling will largely influence the illumination produced by a given arrangement of light sources. The question, therefore, as to whether or not a certain area is properly lighted cannot be decided so accurately from the number and grouping of luminous sources and their candle-power as it can from a method of actually measuring the quantity of light received per unit of area by a particular surface or surfaces whose illumination it is desired to know. In other words, the question at issue is not so much the photometric intensity of the sources of light and their distribution as it is the degree of illumination actually produced, and this question cannot be decided by the photometer, but requires the use of an illumination measurer or illuminometer.

Suppose, for example, it is desired to illumine a billiard room. Though experience must determine the number and position of the lamps, yet the commercial question as to whether or not the necessary illumination has been obtained upon the tables can best be determined, not by the number of separate lamps of a given candle-power, but by an actual measurement of the light received per unit of area on the table

surfaces. In the same way, whether or not a certain grouping of lights over the tables is sufficient to illumine the rest of the room satisfactorily, can be determined by the measurement of the illumination. produced in different parts of the room. Similarly, in a dining or reading room, it is the illumination of the tables that has to be specially provided for, and the question of proper lighting is best determined from a determination of the illumination of the tables irrespective of the character or number of the lights present, except, of course, in so far as æsthetic requirements are concerned.

In street lighting, the requirements are that sufficient illumination be obtained on the surfaces of the streets, houses, vehicles and persons to be able to discern them clearly at reasonably sufficient distances. This primarily requires a certain intensity of illumination over a given area, and can best be determined from the actual measurement of the minimum illumination, rather than by calculations dependent upon a number of complex conditions, of which the photometrical intensity of the sources employed is but one.

Measuring the illumination rather than measuring the intensity of separate lamps would remove a source of great annoyance, both to electrical engineers and to municipalities, by rendering unnecessary discussions as to the candle-power of incandescent or arc lights employed; since, if the contract between the city and the lighting company be so drawn that the illumination in any part of a lighted street, for example, in the absence of moonlight shall not be less than a certain minimum, all questions as to the candle-power of the separate sources will fail to affect the contract, and it only remains for the electrical engineer to provide such lamps as will, at the spacing employed, give the illumination contracted for.

So far as we know, there is no actual apparatus in existence, other than that we are about to describe, that will directly measure the intensity of illumination at a given point, except the portable photometer of Professor Weber, described by Mr. Heinrich at page 422, Volume XI, No. 7, July, 1894, of the proceedings of the American Institute of Electrical Engineers. In Weber's photometer, the illumination received upon a screen from a small standard amyl-acetate or benzene lamp is compared with the illumination to be measured. The difficulty of obtaining a reliable portable standard lamp, the rather cumbersome size of the apparatus and the fact that it does not obviate the difficulty of comparing illuminations of different colors, have led us to devise a new instrument based on entirely different principles. We call our instrument an illuminometer and have filed applications for a patent on the same.

A certain intensity of illumination is required to render a definite object, viewed at a definite distance, clearly delineated to the eye. It is well known that the illumination received upon a printed page of a book or newspaper must have a definite value in order to render the printed characters legible, and that the intensity of illumination so required will, for a normal eye, depend upon the size and character of the print.

We employ the principle in the operation of our illuminometer as follows: A small test object, of, say, printed characters, is placed in a darkened box, sufficiently small to be readily carried in the pocket. The test object is exposed to illumination received from a translucent plate of porcelain or opal glass, which receives directly on its surface the illumination whose intensity is to be measured. Since the test

object receives the light from this plate by transmission and subsequent diffusion, it is the area of the translucent