value of a lubricating oil may be drawn from it and by comparison with other approved oils.

Photometric Tests.

The illuminating power of an oil is determined, with regard to the quantity consumed, by ascertaining the intensity of light emitted by a flame fed with the oil. The intensity of light is determined by a photometer and the quantity of oil consumed by weighing the experimental lamp before and after the experiment.

Photometers are based upon the principle that the intensities of two sources of light, which equally well illuminate different surfaces, are proportional to the squares of their distances.

Of the different photometers in use the most important shall here be described.

Bunsen's photometer. This instrument is almost exclusively used in Germany, partially so in Austria, and with slight modifications in England. In Bunsen's photometer, a vertical sheet of paper with a grease spot at its centre is exposed to the illumination of a standard candle on one side, and the light to be measured on the other. The sheet of paper is placed inside a dark box provided with two plane mirrors placed at such an angle to the paper that an observer can readily see both sides of the paper at the same time.

The box can be slid along a graduated horizontal scale, towards, or from, the light to be measured, and carries with it the standard candle mounted on it at a constant distance of 10 inches. If the box is too near the light to be measured, the grease spot appears brighter on the side of the sheet of paper nearest the candle. If too near the candle it appears brighter on the side of the sheet of paper nearest the light to be measured. The position in which the spot appears equally bright on both sides, is the position in which both sides of the paper are equally illuminated, and the relative intensities of the two lights are then directly as the squares of their distances from the sheet of paper.

Unit of light, or standard flame. All flames used up to the present time are more or less imperfect and do not sufficiently answer the requirements of a unit of light. The flame of a candle or of a carcel lamp is most frequently employed for the measurement of light, though under certain conditions the flame of illuminating gas is also used for the purpose. While in England but one kind of candle is used, which must be of definite composition, in Germany the regulations regarding the candle to be used vary very much and are frequently quite defective. However, candles, as well as carcel lamps, are not constant as regards their intensity of light and subject to variations of greater or smaller magnitude, so that the amyl acetate lamp recently introduced by von Hefner-Alteneck may shortly entirely supplant the candle, it answering best of all sources of light proposed, the requirements of constancy and of being readily produced.

London standard spermaceti candle. The light adopted in the United Kingdom, under the authority of the Legislature, and also in the colonies, as a standard whereby to estimate and express the illuminating value of substances used in artificial illumination, is that which is derived from a spermaceti candle burning at the rate of two grains of spermaceti per minute, or, as usually expressed, "sperm candles of six to the pound, each burning 120 grains per hour." Few candles burn at this exact rate, and in practice, "if the rate of consumption per candle shall not have exceeded 126 grains per hour or fallen short of 114 grains per hour," corrections by calculation are made for such variations from the standard rate. Hence, in England, the consumption of candle-material during the time of observation is also determined, Keates's candle-balance, Fig. 182, being most frequently used for the purpose. The candleholder T, in which the standard candle is secured at the desired height by the screw K, is carried by the knife edge F of the beam A by means of the steel arm L. On the long arm of A is a shiftable weight E, which, when the candle is lighted and burns normally, is slowly shifted towards the centre until

the indicator C, moving in front of the scale D, strikes toward the right. The attachment G on the end of the long arm of A prevents the latter from being raised too much. By the candle burning down its weight is decreased, and the indicator Cagain approaches the zero point of D. When the latter has been reached the time is noted by a watch, and observations regarding the consumption of candle and photometric measurements are commenced. After finishing the latter, the

flame of the candle is extinguished, the time required for the measurements read off, and the consumption of candle during this time determined by placing weights in the pan N until the indicator C points again to zero. The weights used represent the consumption of candle-material during the time of observation. The entire apparatus is secured to one end of the photometer.

In making photometric experiments the flame of the standard candle should be 134 inches long. The standard candles

are 834 inches long from base to shoulder, and according to F. W. Hartley are best prepared for making experiments as follows: Cut each candle in two parts at 43% inches from its base. This is best done by laying the candle on a clean level surface and rolling it under the edge of a sharp knife. Cut away half an inch from the top of the lower half, and also cut away the spermaceti from the top half; reduce the projecting wick of the latter to half an inch in length. Take great care in cutting away the spermaceti, not to injure the wicks in the slightest degree. Burn the pieces of candles not too closely together in a place free from draughts, until the flames attain their maximum size, and each wick bends over and presents a red point through the flame, then touch it with a piece of spermaceti and blow out the flame. Always extinguish candles in the way described, and thus preserve their wicks in good condition.

For conveniently measuring the size of the flame a shiftable. pin is connected with the candle holder, which is provided with two fine metallic points at the desired distance (134 inches) from each other. The lower point is so set as to accurately point at the lowest part of the flame. When the point of the flame touches the upper metallic point, the photometric measurements are commenced. Since it is of importance that the standard candles used should be as uniform in composition as possible, it is advisable to determine the melting or the congealing point of the candle material by the method proposed by Rüdorff. Fill a glass flask of 150 to 200 cubic centimeters capacity two-thirds full of the candle material and melt the latter by immersing the flask in water of from 140° to 176° F. When all is melted take the flask out of the water, dip a sensitive thermometer into the melted material and, while constantly stirring, observe the falling of the temperature. The temperature at which the thermometer remains constant for some time is considered the congealing point; with the spermacetic candle it lies at 43.5 to 44.5° C. (110.3° to 112.1° F.)

Munich stearin candle. This candle is to consume 10.2 to

10.6 grammes (157.41 to 163.58 grains) per hour with a flame of 52 millimeters (2.04 inches), and is to be composed of stearin with 76 to 76.6 per cent. carbon.

Paraffin candle of the Verein der Gas- und Wasserfachmänner. This candle is made of paraffin with a melting point not below 131° F. It is 20 millimeters (0.78 inches) in diameter and of such a length that six candles weigh 500 grammes (17.63 ozs.). The wick should be made as uniformly as possible of 24 cotton threads, and I meter (3 feet 3.37 inches) of it should, in a dry state, weigh 0.668 gramme (10.3 grains). In making measurements the size of the flame is to be 50 millimeters (1.96 inches); trimming the wick of the candle is permissible. According to experiments by Rüdorff, the intensity of light with this candle is not constant, it being in this respect inferior to the spermaceti candle and the Munich stearin candle.

Von Hefner-Alteneck's amyl acetate lamp. Amyl acetate is used for burning in this lamp. The influence of the wick is very slight, and, besides, the size of the flame, after the lamp has burned for a short while, remains constant for a considerable time.

According to von Hefner-Alteneck's proposition the unit of light is the illuminating power of a free burning flame ascending from the cross section of a wick saturated with amyl acetate, which completely fills a German silver wick-tube of 8 millimeters (0.31 inch) internal, and 8.3 millimeters (0.32 inch) external, diameter, and 15 millimeters (0.59 inch) in length. The size of the flame, measured, at least 10 minutes after lighting, from the edge of the wick tube, is to be 40 millimeters (1.57 inches).

A lamp constructed according to these directions is shown in vertical section in Fig. 183 and in ground plan in Fig. 184.

The height of flame is indicated by the line of sight over the two edges a and b. By sighting through the point of the flame to the brightly illuminated edges a and b, and by turning the disk S, the height of the flame is so regulated that the point of the bright nucleus of the flame which appears about 1⁄2 milli