American manufacturers being controlled by standards loaned them by this Bureau.

The manufacturers of high-grade thermometers, including especially those for use at high temperatures, have very generally adopted the methods of annealing recommended by this Bureau as a result of a long series of experiments; hence it is no longer necessary to send abroad for satisfactory instruments of this class. Several manufacturers have installed annealing furnaces and testing plants from plans furnished by this Bureau. An increasing percentage of the thermometers submitted for test require calibration of the highest accuracy, such as calorimetric thermometers and certain grades used in chemical and high-temperature measurements.

Many of the pyrometers tested are the standards used by manufacturers of such instruments. Nearly all American manufacturers and agents of these instruments now have their product controlled by standards furnished by this Bureau. There is beginning to be a satisfactory uniformity in the indications of high-temperature measuring instruments in this country, whereas previously there was a most chaotic condition in this important field.

The Bureau has rendered assistance to manufacturers in the designing and testing of new types of high-temperature instruments and in the development of new methods of measurement. The assistance which the Bureau has been able to render in this way has apparently encouraged several firms to undertake the manufacture and introduction of pyrometers into industrial use.

The establishment and maintenance of correct standards is of fundamental importance. From the nature of the problem no concrete standards can be constructed for temperature measurement and the scale must constantly be reestablished and checked, which involves a large amount of work and time. It is only within the limited range of temperatures of 0° to 100° C. that the Bureau has yet been enabled to maintain a standard temperature scale with adequate accuracy. No work has yet been attempted on the lowtemperature scale, and only a start has been made on the hightemperature scale. Preliminary work has been done toward the establishment of the high-temperature scale to 2,000° C., by means of the laws of radiation; and for the intercomparison of the hightemperature scales as given by electrical-resistance, thermo-electric, optical, and total-radiation methods.

The work in heat measurements has been mostly investigational, or in connection with cases of disagreement between litigants involving unusually extended experimental work. Such heat measurements include tests of inflammatory materials, melting points of metals, behavior at high temperatures of metals, bricks, and other materials, and cooling curves of steel.

In collaboration with the American Society for Testing Materials, exhaustive tests of samples of lubricating oils from various sources, by different methods, have been begun by this division.

Work on the thermal properties of the materials used in the new filament lamps has been undertaken, but is not completed. It was hoped to be able to determine the melting points of some of the very refractory metals both in the form of filaments and in the form of minute particles by a new radiation method; but this work has thus

far been limited to the determination of the melting points of tungsten and tantalum in the form of filaments.

The introduction in the industries of radiation methods of measuring high temperatures requires the determination of the errors of such instruments when examining incandescent substances, such as iron, copper, fire brick, etc. A few preliminary measurements have been made on the radiation from iron. Some experiments have been made on the constancy of melting of certain refractory metals in order to define and construct a new standard of light based on the intensity of light emitted from an inclosure immersed in a metal at its freezing point. A study has been made of the available methods of obtaining cooling curves for the thermal analysis of steels, alloys, and chemical compounds, the results of which have been published in the Bulletin.

The determination of the specific heats of brine solutions, data of great value for the refrigerating industries, has been nearly completed and a preliminary report published. A new type of vacuum jacket calorimeter suitable for the determination of specific heats at low temperatures has been developed, and several very sensitive resistance thermometers have been designed and constructed for calorimetric work. It is hoped that the work in calorimetry may be greatly extended, and certain gas calorimeters have been acquired with a view to an examination of their behavior.

The apparatus is being assembled for repeating and extending the work of Joule and Thomson on the free expansion of gases, a problem fundamental to the conception and measurement of temperature.

There were tested during the year 9,193 thermometers, including 202 high-temperature thermometers, 536 laboratory and special thermometers, 60 low-precision industrial thermometers, and 8,395 clinical thermometers; 9 optical pyrometers, 41 thermo-electric pyrometers, 12 resistance pyrometers; 99 viscosity and flash-point tests of oils; and 17 other tests involving heat measurements; in all about 9,351 separate tests. There was a considerable increase over the preceding year in the testing of high-grade mercurial thermometers and in oil tests for the United States Government.

DIVISION IV.-OPTICS.

SPECTROSCOPY.

Important investigations have been conducted in this section of the Bureau during the year. Work on the helium tube as a possible primary light standard has been continued. Age tests and the effect of varying the glass of the stem remain to be carried out. The purity of spectral light sources has been investigated further, recent work being devoted to the structure of composite lines. The intensity of any spectrum line was shown to be probably expressible in terms of but a single variable and two constants whatever the form of excitation of luminescence. The relation of light to radiation and visual sensibility has been examined in more detail, and practical deductions drawn.

The structure and properties of compound lenses have been studied, particularly the effect of the position of the transverse axis on astigmatism and distortion. Preliminary specifications suitable for de

fining lens "speed" were drawn up and applied to a photographic lens with a Martin's photometer.

A number of photographic lenses were tested as to fidelity of reproduction. The transparency for visible and ultra-violet radiation of a special absorption cell and of several samples of glass have been examined. The refractive indices of two samples of glass and a fluid were determined through the visible and ultra-violet portions of the spectrum. A special design of spectrophotometer was constructed and tested. A Lummer-Gehrcke plate and special camera for the echelon have been added to the spectroscopic equipment, and a precision optical bench for lens testing is being designed and constructed.

RADIOMETRY.

The measurement of radiant energy is still in a primitive state of development, in particular its measurement in absolute units. This is chiefly due to the fact that the radiation to be measured is generally from a surface, the temperature of which is practically indeterminate, as is also the energy expended within the radiating substance. The measurement of the energy radiated is exceedingly difficult, but progress is being made in the development of apparatus and methods for these measurements.

The recent development of processes which require an accurate knowledge of temperatures necessitates a study of the laws of radiation of various materials with variation in temperature. This is notably true of metals used in incandescent lamps to ascertain the cause of their luminous efficiency. Three investigations of such materials have just been completed. An investigation of the Nernst glower radiation shows that its emission spectrum is discontinuous at low temperatures, gradually becoming continuous at high temperatures. No "radiation constant" was found, so that the results of previous experimenters are unreliable. An investigation of the radiation of solids in the form of electrically heated rods, and fine powder brought to incandescence on a heater, show the spectral partition of energy which occurs in sharp emission bands, superposed upon a weak continuous spectrum. The maximum emission moves toward the short wave lengths with temperature rise; but the bands remain fixed in position, simply increasing in intensity-a result entirely unexpected from previous observations in the visible spectrum.

The most important constant which occurs in the spectral radiation formula of metals, such as platinum, osmium, tungsten, tantalum, etc., in the form of filaments in incandescent lamps, has been investigated. This particular radiation "constant" also occurs in Stefan's law of total radiation, where it is given as 4 for a complete radiator. This important so-called constant was, however, found to decrease with rise in temperature, with a value higher than platinum at any given temperature. This accounts for the high luminous efficiency of metal filament lamps, and incidentally verifies the theory which requires the optical constants of the metals to be a function of the temperature and the wave length. This seems to have escaped the attention of previous experimenters.

Numerous inquiries concerning radiometers, the transparency of gases for heat waves, and of absorption screens for infra-red work have been received and appropriate information given.

POLARIMETRY.

Polarized light is utilized in testing all sugars bought and sold or imported. The polariscope and its accessories are therefore the indispensible adjuncts of the sugar refiner, the beet and cane sugar planter, and sugar laboratories, but such apparatus has remained unstandardized. Hence it is futile to expect tests by various interests to agree when not only apparatus but methods of testing differ so greatly. After extended investigation of the theory and practice of polarimetry, the basis of standardization of polarimetric apparatus is now rigidly defined by this Bureau. During the year a number of quartz plates and 1,270 samples of sugar for the Treasury Department were tested by this section.

The most important piece of work of the year was the successful development of a quartz compensating polariscope with adjustable sensibility. The theory and construction of this instrument are described in the Bulletin, and it has been adopted by the Secretary of the Treasury for use in the customs service.

The disturbing effect of atmospheric conditions on raw sugar tests is appreciable, as an investigation begun over a year ago clearly shows. This work will be continued until satisfactory correction has been obtained.

Large quantities of pure sugar are indispensable for research work in polarimetry, and numerous requests from various sources have been received during the past year for such sugar for calorimetric and other purposes. To meet these demands apparatus for preparing pure sugar by recrystallization has been developed. The Bureau will soon be able to produce standard sucrose in large quantities under controlled conditions which guarantee a high degree of purity.

A new optical system giving monochromatic light when used with intense sources has been developed for use in polariscopic work, and a compact form of apparatus is being designed.

DIVISION V.-CHEMISTRY.

The chemical division has been actively assisting government departments as well as the Printing Investigation Commission of Congress. The investigations commenced or continued on behalf of the departments and the Government Printing Office, and in which more or less progress has been made, relate to the testing of materials in connection with contracts and purchases. The work is of importance, not only to ascertain whether supplies furnished by contractors conform to specifications, but also as a preliminary to the formulation of specifications for future use where none is now insisted on. There is a wide field of chemical usefulness open to the Bureau in lines of research relating to the purchase of government supplies. The researches aiming at improvement in the quality of and methods of testing these materials have hardly more than begun, and will consume much time if results of real value are to be expected. In order to expedite this work and permit such routine testing as will be called for, a considerable addition to the working force and laboratory space is imperative.

The classes of problems planned and under investigation for the Government are numerous, and in all about 1,766 tests and analyses have been reported in the following classes during the year:

(1) Testing of writing, printing, copying, and carbon papers; writing, copying, record, and typewriter inks; typewriter ribbons; mucilage and sealing wax, and researches bearing on the preparation of suitable specifications, with a view of improving the quality, reducing the cost, and rendering the Government more independent of manufacturers than at present. There have been reported by the Bureau about 1,595 tests and analyses of papers, inks and ink powders, printer's ink, and mucilage.

(2) Tests and analyses of lubricating oils, linseed oils and driers, paints and varnishes, of which 139 separate tests have been made during the year.

(3) Miscellaneous tests of soaps, rubbers, metals and alloys, and other materials, about 32 in number.

(4) Analyses of standard steels intended for distribution to all who wish to test the accuracy of their own chemical work and that of their employees. The increasing demand on the part of those engaged in the iron and steel industries, and of analysts and others, for the standard irons and steels, we have been preparing with the assistance of the American Foundrymen's Association and the Association of American Steel Manufacturers, and the continued expression of interest in the subject, shows the value of this line of work. Of the irons and steels thus far prepared 465 separate samples were supplied to applicants during the year.

Further expansion is needed to meet the demands for cooperation on the part of other divisions of the Bureau. Cooperative problems pressing for solution that can not be solved without the aid of chemistry are: (1) The preparation and testing of pure materials used in the construction of standards of electromotive work; (2) the preparation of many substances of the very highest degree of purity, and the determination of certain constants of these substances for use as standards in calorimetric work; (3) the determination of the physical properties of metals and alloys used in the arts demands a detailed knowledge of their chemical composition. The successful carrying out of these researches will be of very great value to the wide range of industries in which these problems are of vital importance.

ENGINEERING INSTRUMENTS.

The demand of government departments and the public for information regarding engineering tests has greatly increased, and many instruments have been submitted for test, including anemometers, speedometers, meter provers, water meters, paper testers, pressure gauges, steam valves, and the work was mainly for practicing engineers or municipalities. A small mercury manometer, with maximum capacity of 60 pounds per square inch, has been erected for the calibration of gauges of low range, and one of greater capacity will be installed in the new building.

A study has been made of the effects of air drag, surrounding walls, wind currents, and centrifugal forces upon the present method of