arts by an experienced eye. The temperature scale of the bureau is now reproducible to about 1° at 1000° C. and to about 10° at 2000° C.
It should not be supposed that the constant efforts for higher accuracy are made to satisfy the dreams of the pure scientist; they are demanded by the requirements of technical tests and commercial processes. The character of some products is materially affected by a variation of 20° in the very high temperatures employed in their manufacture, and an accuracy of about 10° is required. To meet this demand there must be the exercise of man}*-precautions and the elimination of many sources of error.
The importance of great accuracy is well illustrated by a test which the bureau was called on to make in a dispute between a purchaser and a seller of coal, in a case where the contract was based on the heat value of the coal, with a penalty clause for any deficiency in the heat value, and a premium for any excess above the stipulated one. A difference of about 0.05° was found between the thermometers used by the two parties. While this difference was small, it was sufficient to bring the parties into agreement, and to make a difference of some $25,000 a year in the money paid for the coal.
The division of optics has been engaged in many investigations of moment, but none of more practical value than the improvements in the application of polarized light to the testing of sugar solutions by means of the polariscope. Plane polarized light differs from common light in having all its vibrations reduced to a single plane. The optical property of a sugar solution utilized to determine the amount of sugar present is its property of rotating the plane of polarization when the polarized light passes through it. The degree of rotation determines the per cent, of sugar present.
The bureau is able to make immediate application of its research work in this field, thus directing public attention to the results attained. As a result of the polariscopic tests of imported sugars at the bureau, the differences in the findings at the five principal sugar ports of entry have been reduced to 0.2 per cent. The importance of this work grows out of the fact that it increases the accuracy of the tests made on dutiable sugars and rigidly defines the scientific basis on which the revenues from them are collected.
The division of chemistry is in a large way auxiliary to all other divisions and cooperates with them in giving such service as chemistry alone can offer. It has been indispensable in the work of the division of electricity; it has prepared materials in the purest form for setting up Weston normal cells as standards of electromotive force, and for use in the silver voltameter for the international unit of electric current.
Much of the labor in testing supplies offered by the bidders under the new system of purchase by the General Supply Committee devolves on this division. The analyses of writing and printing inks, paper, and