For

area either above or at one side, and the illumination from such area does not reach back into the room far enough, prisms and diffusing glasses of various kinds are applicable. The action of the prism glass window is to bend the light rays so that they strike back further into the room than if a clear glass window were used. Rough and ribbed glasses accomplish the same end with less precision and effectiveness. They diffuse the light rays passing through, and a certain portion of such rays are directed back into the room. some locations louvers or shutters consisting of partially or wholly opaque strips which can be tilted at any angle make it possible to regulate the relative amount of sun and skylight or cut out direct sunlight without too serious a reduction in the skylight. The common method of controlling sunlight is by translucent shades but this method for some interiors (such as art galleries) does not offer very accurate control.

The Brightness or Intrinsic Brilliancy of Various Artificial Light Sources and also the brightness of some sources equipped with diffusing glassware for the protection of the eyes is shown in Table IV.

Depreciation due to dirt on glass and reflecting surfaces and to inherent characteristics of the lamps used must be recognized in design.

Both the total lumens and the lumens per watt of tungsten filament electric lamps drop with use, partly by the blackening inside the bulb and partly by disintegration and increase in resistance of the filament. Such lamps operated at the specific outputs shown in Table II, fall off in lumens output about 15 per cent. in 1000 hours service. With electric arc lamps and gas mantle burners so much

depends upon the adjustment and other variable factors that no depreciation figure inherent in the lamp can be given, but unless maintenance is especially good more must be allowed than for the internal depreciation of the tungsten filament electric lamp.

The accumulation of dirt on the surrounding glassware and on the globe or reflector is an important cause of loss of light and should also always be reckoned with in preliminary calculations. It is necessary to assume some probable maximum depreciation figure from this cause and in making such an assumption of course the surrounding conditions and the probable frequency of cleaning must be considered. In Table V is given a compilation of results of various tests made in different places by different observers on the effect of the accumulation of dirt, and Fig. 2 shows the depreciation over an extended period for a given set of reflectors.

The effect of accumulation of dirt on side and ceiling windows is probably about the same as on lamps.

Utilization of the Generated Light Flux.-There are various methods of calculating the resultant illumination at the desired point with a

TABLE V.-LOSS OF LIGHT BY ACCUMULATION OF DIRT

NOTE. Since depreciation is more rapid at first, as shown by the curves the decline per month here given would not apply to longer periods.

NOTE. For extensive additional tests see paper by A. L. Eustice, Trans. I. E. S., 1909, p. 849.

given generated light flux. Before making such calculations it is of course important to reach an intelligent decision as to the points where the desired illumination is needed and whether it is best to consider the illumination measured in a horizontal plane, or vertical plane, or a plane in some other angle, suited to the particular requirements in question. Common practice in calculating and measuring the illumination in most interiors is to ascertain the illumination in a horizontal plane from 2.5 to 3.5 ft. above the floor, or about the height of desks, counters and benches. For the

majority of interiors this consideration of the horizontal plane serves the purpose sufficiently except for special localized lighting around machinery. If the illumination in the horizontal plane, commonly known as the "working plane" is to be taken as the criterion, it is possible to measure the average illumination in this plane over an entire room by measuring the illumination with a portable photometer at the center of a number of equal-sized rectangles into which the room may be divided. Dividing this average light flux by the light flux generated by the lamp gives what is known as the percentage efficiency of utilization, or utilization factor. Of course any other plane might be used for figuring efficiency of utilization provided the position of the plane were the position where the light was wanted. For example in an Art Gallery the efficiency of utilization might well be figured from the light flux incident upon wall spaces devoted to pictures and in a show window it would be figured from the flux through a curved surface corresponding to the line of trim of the window.

The point-by-point method of calculation (that is, if dealing in English units, dividing the candle-power by the square of the distance in feet to the point in question and multiplying this by the cosine of the angle of the incident ray to the surface in question to get the foot-candles incident illumination) is now chiefly used only for calculating the illumination at a few points from a single or small number of light sources. It is too time-consuming and laborious a method for the calculation of the illumination of large interiors with many light sources. It has the further limitation that it takes no account of reflection from ceiling, walls and floors and considers only the illumination direct from the lamp and its accessories.

The point-by-point method may be of considerable use in forecasting the differences in daylight illumination and at different points of interiors where the sky exposure and reflection coefficient of the buildings visible from any point in question are definitely known. The foot-candles illumination at various points as one proceeds back into a room from a window with unobstructed sky exposure may for the rough purpose of practical calculations be taken as inversely proportional to the square of the distance from the window to the given point. In applying this rule the fact should be kept in mind that frequently the window is far from an unobstructed sky exposure and that the sky exposure changes as seen from various points further back into the room. The effective exposure is the projected area of the sky seen by one looking at the window from the given point.

A practical short-cut in the use of the point-by-point method in calculating horizontal illumination which obviates the necessity of a table of cosines and makes possible calculations with only the aid of a polar candle-power curve of the light sources, is the following, which is a graphic method of applying the cosine factor. In the usual rule for getting horizontal illumination the illumination is equal to the candle-power at the given angle divided by the square of the distance multiplied by the cosine of the angle between the ray in question and the vertical. Now if we draw a perpendicular from the photometric curve at the angle in question to the vertical and take as the candle-power the candle-power scale reading at the point where this perpendicular intersects the vertical, we apply the cosine factor at the outset and by simply dividing this candle-power at the intersection with the horizontal, by the square of the distance the illumination is determined.

In calculations of illumination by the zone flux method all of the lumens emitted in a certain zone, say from o to 60 degrees or from o to 70 degrees, are figured as falling upon the working plane in the general lighting of an interior. This method, of course, takes no account of the uniformity of illumination and where approximate uniformity is desired must be used only with lamps and reflectors giving a type of distribution which will be sufficiently uniform. The zone flux method is chiefly applicable to illumination calculations with opaque reflectors where all of the flux is emitted in downward directions and little reliance is placed upon walls and ceilings to bring up the general illumination. Some industrial plants and foundries present such conditions. In the application of this method care must be taken not to select such a large zone as a basis that too much of the light strikes walls or other obstructions. At the same time in large interiors it is not necessary to confine the zone to simply those which would cover the floor near by. In show-window lighting if the reflector selected is such as to confine its flux to the plane it is desired to illuminate the method may sometimes be used for approximation.

Empirical methods of calculation based on actual experience and tests of existing installations form by far the most important basis for most calculations. With the other methods certain assumptions are necessary which may or may not be correct. With the empirical method based on experience, the only sources of error are those due to erroneously assuming conditions in the case to be calculated to be similar to those in the tested cases. Tables VI and VII and Figs. 3 to 9 inclusive give utilization factors or ratio of generated lumens to