the need for such work, and an increased demand for tests of flame standards has resulted. The Bureau certified a larger number of such standards than in all previous years combined. Each test provides a standard by which the entire gas supply of a city may be maintained at standard lighting power.

The reference color standards used in the inspection of cotton oil are being standardized. Extended comparisons of direct midday sunlight with the acetylene flame has shown that sunlight is a satisfactory constant working standard for an ideal white light.

Standard magnetic test bars have been loaned to educational institutions to enable them to check their magnetic measuring instruments. This will promote uniformity in such measurements. The cooperation of national technical societies has been secured for both experimental work and the consideration of specifications for magnetic standards and tests.

The Bureau must steadily extend its comparisons to include standards for which facilities and assistants are still lacking. The systematic comparison of industrial standards with the fundamental standards will greatly aid industrial development.

CONSTRUCTION OF STANDARDS.

It is necessary that the Bureau construct standards whenever improved methods are ascertained or developed at the Bureau or as standards deteriorate. In some cases standards have not heretofore been available and must be developed. This usually involves scientific investigations of the most advanced character, since the perfection of such standards should be in advance of the technical requirements in science and in industry.

Standards of electric resistance and electromotive force have been investigated to make them more trustworthy. The silver voltameter investigation to fix the primary unit of current is completed and the results are now in press. The electric balance, by which the electric current is weighed and the unit of current independently ascertained, has been perfected to a high degree of accuracy and the completed results of an extended determination are being prepared for publication. These researches upon the fundamental electrical units and standards involve the utmost instrumental refinement, the development of the rigorous methods of measurement, and painstaking study of the theory of such measurements, as well as extended experimental work covering several years. The completion of these researches marks a notable advance in electrical measurements, upon which the progress of electrical science and industry so largely depends.

A number of the standards of electric resistance of the sealed type devised at the Bureau are sent each year to the several national

laboratories of Europe. The comparisons of these standards show a high degree of concordance, the difference not exceeding 1 part in 30,000. The close agreement is due largely to the Bureau's method of sealing such standards to avoid the effects of humidity. The Bureau prepares standards of light-incandescent electric lamp standards (both carbon and tungsten) of certified illuminating value, and also standard color screens. Through standardized lamps the primary unit of light of the Bureau is maintained. Such standardized lamps are also sold to the public.

Vacuum tubes of improved design filled with pure rare gases have been supplied in increasing numbers to various optical laboratories for use as standard light sources for technical purposes. These tubes have developed remarkable practical uses for the optical properties of the inert components of the atmosphere. A great number of card and glass reference standards of color have been standardized and certified. Methods of color analyses applicable to industrial products have been developed.

Standards of composition for various irons, steels, and other materials are prepared, analyzed, and certified by the Bureau. These are distributed to the public for a small fee and serve to check the correctness of the methods of industrial analyses and the skill of the analysts. More than 30 accurately analyzed materials are now available for distribution. Three new standards of special steels, namely, nickel, chrome-nickel, and chrome-vanadium steel, were prepared for distribution during the year. Other standard materials are in preparation, including sodium oxalate, for use as a standard for oxidimetry, and benzoic acid as a standard for acidimetry. During the year 168 standards of combustion were prepared for chemists and engineers for standardizing calorimeters and to check the accuracy of the methods of determining the heating values of fuels. The materials include pure sugar, napthalene, and benzoic acid of exceptional purity and ascertained heating value, certified in each case by the Bureau.

TESTING OF STANDARD MEASURING APPARATUS.

During the past year a great variety of standard measuring apparatus has been compared and, in many cases, new methods of measurement have been developed of greatly increased precision. It is essential that the Bureau keep in advance of the needs of science and industry in rigor of methods and precision of results. Among the important work of the year may be mentioned the following:

The experimental study of the properties of electric condensers has been continued with excellent results. From a detailed study of the theory of the moving-coil galvanometer it has been found possible, after five separate measurements, to compute the constants and per

formance of such instruments accurately. These data will greatly aid in the selection of suitable instruments for special purposes. Improved methods and apparatus for magnetic measurements have been devised for testing specimens larger or smaller than standard sizes, greatly enhancing the range and usefulness of this work. The Bureau has designed, constructed, and tested new apparatus for testing electric potentiometers. A technical study of the theory of the corrections to such apparatus has been completed which will place this work upon a more exact basis.

More than 3,000 tests of standard weights were made for the National and State Governments and for technical laboratories. The standardization of fine-mesh sieves used in testing cement was thoroughly investigated, new specifications developed for the Government, and new and accurate methods of measuring the mesh of such sieves devised by the Bureau.

The field investigation of commercial weights and measures of the United States was concluded during the year. The investigations covered every State in the Union and was made by experienced inspectors, and was the first general investigation of the weights and measures throughout the United States. The results were furnished in manuscript to the States where pending legislation called for a study of the needs for adequate inspection service. Already the investigation has resulted in the passage of many State laws upon the subject. The Bureau cooperates with the States in framing the text of the technical details of such legislation, though the Bureau force is inadequate to meet the demand. This cooperation has been appreciated by the States. The investigation is being summarized in a report soon to be published. A startling proportion of the measures of daily trade were found incorrect, e. g., scales, 44.53 per cent; weights, 15.9 per cent; dry measures, 48.5 per cent; and liquid measures, 28.2 per cent.

A notable improvement is now found in American-made glass volumetric measuring apparatus submitted to the Bureau for test. The rigid specifications of the Bureau have done much to raise the standard of such volumetric apparatus until it is now superior in some respects to the best foreign product. Practically all volume and density measuring apparatus used in Government laboratories is now ordered on Bureau of Standards specifications and tested by the Bureau. More than 9,000 capacity measures were tested during the year. Aneroid barometers, for altitude indications, were studied in great detail during the year, as to their behavior under mechanical test, temperature variation, atmospheric conditions, and air-pump tests. Elaborate observations were made as to each factor involved. As a result, the Bureau will undertake at the request of Government departments and of manufacturers the development of a standard

specification for aneroid barometers. Methods and apparatus were developed for precision testing of mercurial barometers and with a rapidity consistent with satisfactory results.

Nearly 18,000 thermometers were tested by the Bureau during the year. These include standards used to determine the heating value of fuels, primary standards used to define the standard temperature scale, mercurial thermometers, clinical thermometers used by the medical profession, electrical resistance thermometers, and thermocouples. A gas calorimeter test was made for public-service commission use. Such work will increase in importance in view of the changing legal requirements for the sale of gas. The heat tests also included the determination of specific heats of special distillates, the heats of combustion of engine oil, and a by-product of oil gas. The testing of high-temperature measuring instruments increased about 60 per cent over that of the preceding year. These comprised platinum resistance thermometers, pyrometer galvanometers, and optical and radiation pyrometers.

A new, rapid, and precise instrument for determining reflecting power has been perfected, used in testing at the Bureau, and made available to the public. Many quartz plates for the daily control of saccharimeters used in sugar testing have been certified for the use of the Government and the general public. More than 1,300 polarimetric and other optical tests were made during the year.

Radiometry is one of the newer sciences which are immediately applicable industrially. A bulletin was prepared giving the advances in the design and construction of radiometric apparatus. The many inquiries which were being received from specialists and others made such a paper especially timely.

PROBLEMS CONNECTED WITH STANDARDS.

Extended tests of certain classes of electrical measuring instruments were made by the Bureau for the makers to show where improvements were needed. Such comparisons go far to advance the art of measurement and the results are of great value to manufacturers. A new method has been devised for testing electric transformers and a new apparatus has been designed and constructed to measure time in the testing of watthour meters by an apparatus quicker than the chronograph and more reliable than the stop watch. The measurement of light involves practically every branch of the new profession of illuminating engineering. The range of problems connected with standards of color, distribution, reflecting power, intensity, efficiency, etc., are practically all problems of measurement for the effective solution of which access to standards are necessary and elaborate equipment needed. New methods of measurement are developed to govern each new type of lamp. The solution of such

problems is of vital concern in view of the close connection of proper lighting with human well-being.

The measurement and control of high temperatures is receiving rapidly increasing attention in many industrial lines, and as in other lines the Bureau laboratories are visited almost daily by technical men interested in such problems. A study of the radiation from surfaces of steel at high temperatures is being made. This work is necessary in order to enable observers using radiation pyrometers to obtain true temperatures, a matter of the utmost importance to efficiency in the metallurgical industries. Two problems of importance were studied and the results published, one relating to radiation from a standard of radiation technically known as a "black body" and the other upon the steam expansion line in steam turbines.

The solution of the problem of stem corrections of mercurial thermometers was completed during the year and the published results form an important contribution to the literature of precision measurement in this field. A theoretical and experimental investigation of the effects of thermometric lag on temperature measurements was also completed. These researches facilitate the standardization of thermometers with high precision.

The Seventh Annual Conference on the Weights and Measures of the United States was held at the Bureau in February, 1912, for the consideration of the technical problems in connection with the administration of weights and measures inspection service. Twentyfive States and 51 cities were represented, and the conference was the most successful thus far held. Through this conference the Bureau is leading the movement for correct weights and measures in daily trade, a movement which has now become nation wide.

The Bureau's investigation of the 100 per cent point of the saccharimeter has brought to light an important error in the previously accepted value of this fundamental constant. The application of polarimetry to the standardization of sugar analysis by physical methods has been continued. The Bureau has controlled the accuracy of the sugar analyses at the various ports by check analyses of sugar samples received daily. The type of saccharimeter as developed at the Bureau is now in use at the various customhouses, as are the methods prescribed for such work.

To measure radiant energy, whether light or heat, a precise knowledge is needed of certain constants or data with which to compute temperatures beyond ordinary ranges and for other purposes. The Bureau has for some years been engaged upon the determination of the constants of radiation, a work requiring elaborate experiments, study, and computation. Such work underlies scientific advance in several important fields, such as pyrometry, photometry, and spectros