The Estimation of High Temperatures by the Method of Color Identity. C. C. PATERSON and B. P. DUDDING. (Proceedings of the Physical Society of London, vol. xxvii, part iii.)-(1) Preliminary experiments are described on the method of “color identity” adapted to the estimation of the temperature of incandescent substances, such as metal or carbon radiating in the open; by this method the “true” temperature of certain bodies as distinct from their “ black body” temperatures can be arrived at with a very fair degree of accuracy. (2) By the color identity method the total luminous radiation (white light) from a black body is made identical in color with that from the incandescent metal under examination by adjusting the black body until there is color identity in the field of a Lummer-Brodhun photometer. (3) Comparisons are made with the results so obtained with those obtained by other methods, and the color identity method is shown to give the correct result for melting platinum. (4) Formulas are deduced, based on the fundamental theories of energy radiation and the sensitivity of the eye, connecting the temperature of carbon and tungsten filaments with their lumens per watt, and it is shown that these expressions hold from the lowest to the highest values of lumens per watt. (5) It is shown that the color identity method of determining filament temperatures is practically independent of the cooling at the ends of the filaments of ordinary lamps. (6) An explanation is given of the principal factors and limitations of the color identity method, in which it is shown that accurate results should be obtained so long as the bodies under consideration act as “gray” bodies throughout the visible spectrum, and that there will be a tendency to error to the extent that they depart from the gray body condition in the visible spectrum. (7) The color of the radiation from melting platinum is shown to be the same as that from a carbon filament lamp operating at 2.65 lumens per watt or 4.75 watts per mean spherical candle or, approximately, 3.8 watts per mean horizontal candle. Largest Commercial Gasoline Engine Ever Built. Ayon. (Scientific American, vol. cxii, No. 24, 588.)—What is said to be the largest gasoline engine ever built has been installed in a doubleended ferryboat utilized for the transportation of trains across an arm of San Francisco Bay. Although designed for marine work, it shows to a remarkable degree the influence of automobile practice in its general appearance. Weighing approximately 120,000 pounds, this 600-horse-power unit is the climax of a gradual development of the heavy gasoline engine on the Pacific coast. It has four cylinders of 16-inch bore, each measuring nearly six feet and weighing 1700 pounds. The normal speed of the engine is 225 revolutions per minute. The trips are not long, and between runs all fuel consumption ceases, effecting a considerable saving over steam despite the high cost of fuel for operating an engine of this size. NOTES FROM THE U. S. BUREAU OF STANDARDS.* A STUDY OF THE QUALITY OF PLATINUM WARE. [ABSTRACT.] THERE has been devised a simple, thermo-electric method. suitable for the determination of the purity of platinum ware. This method does not mar the article tested, and gives data for the classification of platinum in terms of its equivalent iridium (or rhodium) content. There were examined by the thermo-electric method 164 pieces of platinum ware, of which 26 per cent, contained less than 0.5 per cent. iridium and 67 per cent. less than 2 per cent. of iridium. Of 84 crucibles, 36 per cent. contained less than 0.5 per cent. iridium and 87 per cent. less than 2 per cent. iridium. A method has been developed for determination of the exact loss on heating of platinum crucibles by means of a suitable electric furnace containing no heated metal parts. Fourteen crucibles of various makes and grades were examined for loss in weight on heating and after acid treatment following each heating. Their magnetic susceptibilities were also determined. The susceptibility of pure platinum is zero, and the range of susceptibility of seven crucibles is i to 125. The heating losses per 100 cm.? of practically iron-free crucible surface at 1200° C. ranged from 0.71 mg. to 2.69 mg. per hour, the lesser losses being for crucibles containing rhodium and the greater losses being associated with iridium. Iron appears to lessen somewhat the loss of weight on heating, but its presence is objectionable on account of the soluble oxide formed on the crucible surface. The chemical analysis and magnetic measurements place the crucibles in only approximately the same order as to iron content; the magnetic susceptibility is not, however, proportional to the iron content. * Communicated by the Bureau. † To appear as a Scientific Paper of the Bureau of Standards. It appears to be possible, from thermo-electric and microscopic examinations of a crucible, to predict its probable loss of weight on heating within limits close enough for analytical purposes. Suggestions are offered concerning the specifications of highest grade platinum crucibles, including the substitution of rhodium to 5 per cent. for iridium, and the practical elimination of iron. Whether crucibles have been long in use or not, after the first two or three heatings and acid washings, appears to make little or no difference on their behavior as to losses on heating and washing. The nature of the process of disintegration of platinum and its alloys is briefly discussed. A DIRECT-READING DEVICE FOR USE IN COMPUTING CHARACTERISTICS OF VACUUM TUNGSTEN LAMPS. By J. F. Skogland. [ABSTRACT.] It has been shown in a previous paper * that the ordinary characteristic relations of vacuum tungsten lamps may be expressed with high precision by a set of characteristic equations, each involving two variables; or solutions may be made more quickly by employing tables computed from the equations. For example, having given observed values of voltage, candle-power, and watts per candle, the values of candle-power and watts per candle at any other voltage are obtained from the equations or tables as follows: 1. From the observed values of voltage and candle-power their normal values,—that is, their values at normal watts per candle (1.20),—are computed. 2. The ratio of the desired voltage to the normal voltage just found is computed. 3. Substitution of this voltage ratio in the proper equations, or reference to the corresponding point in the tables, gives a candle-power factor and the actual watts per candle. * Middlekauff and Skogland, “Characteristic Equations of Tungsten Filament Lamps and Their Application in Heterochromatic Photometry," Bulletin of the Bureau 11, p. 483 (1914); Scientific Paper No. 238. 4. The normal candle-power is multiplied by the candlepower factor to obtain the desired candle-power. This process, though simple, requires considerable time before a solution is obtained. The direct reading device here described and presented ready for use solves directly and without preliminary reduction all problems introducing as variables the voltage, candle-power, and watts per candle of vacuum tungsten lamps. Its construction depends directly upon the characteristic equations mentioned above. This device consists of volt, watts per candle, and per cent. candle-power scales. The watts per candle and per cent. candlepower scales are fixed in their relation to each other. The volt scale is arranged to be detached from the plate, so that it may be applied to the fixed scales at the point corresponding to observed values of watts per candle and voltage. From a single setting of the volt scale to observed values within the range of from 0.70 to 2.05 watts per candle may be read values of any one of the variables at any other point within this range. The volt scale has a range of from 94 to 166 volts. The chosen limits of watts per candle and voltage are considered sufficient to include settings and solutions for 105- to 130-volt lamps in standardizing and life test-work. A test of the device, illustrated in part by examples, indicates that, on an average, values of per cent. candle-power and of watts per candle read therefrom deviate from those obtained by use of Tables 20 and 22 of the paper just referred to by amounts not exceeding 0.10 per cent, and 0.05 per cent. respectively; also that values read from the scales check observed values very nearly as closely as those computed by use of the tables. All of the points (referred to horizontal and vertical scales of equal parts) through which lines of the scales of this device were drawn have been tabulated, so that similar scales of the same or different range may be constructed directly from these values. Sufficient discussion of the derivation of these values and of the relative position of the scales has been given to direct the construction of scales not included within the watts per candle range here employed. By the general method of this paper any related functions of exponential form and of the same degree as the characteristic equations may be represented and used as scales of a similar device. VOL. CLXXX, No. 1075-8 The chief merits of the device when compared with other methods of characteristic evaluation seem to be its simplicity, precision, and saving of time resulting from its use. Solutions are made directly, without reference to normal watts per candle, voltage ratios, exponents, etc. These considerations should recommend it to testing and standardizing laboratories. CONSTRUCTION OF PRIMARY MERCURIAL RESISTANCE STANDARDS. By F. A. Wolff, M. P. Shoemaker, and C. A. Briggs. [ABSTRACT.] The paper deals with the construction of four one-ohm mercury standards of resistance in accordance with specifications adopted by the International Conference on Electrical Units and Standards (London, 1908). The London Conference defined the international ohm as the resistance offered to an unvarying electric current by a column of mercury at the temperature of melting ice, 14.4521 grammes in mass, of a constant cross-sectional area and of a length of 106.300 centimetres. Owing to the impossibility of exact realization of the above conditions, principally because of the impracticability of securing glass tubing of strictly uniform bore, certain specifications were essential. Those adopted specified that the tubes used must be made of a glass the dimensions of which change little with time; that they be well annealed and straight; that the bore be as nearly as possible uniform and circular; that the area of cross-section of the bore be approximately one square millimetre, and that the mercury have a resistance of approximately one ohm. It was also specified that each tube be accurately calibrated, and that no tube have a calibre correction greater than 5 parts in 10,000. The length of the tube, the mass of mercury the tube contains, and the electrical resistance of the mercury must all be determined at a temperature as near to o° C. as possible, all measurements being corrected to o° C. The four tubes used at the Bureau were selected from a large number of tubes of Jena 59 III glass, specially drawn by Schott |