Report of the National Committee on Micrometry.

Dr. Lester Curtis, Secretary pro tem., for the National Committee on Micrometry, presented the following report:

The National Committee on Micrometry would respectfully submit the following report to the American Society of Microscopists:

We would recommend the adoption as a standard by the Society of the bar placed in our possession by Dr. R. H. Ward, Secretary of the committee, and adopt without change the accompanying report of Prof. Hilgard, Superintendent of the United States Coast Survey, and Director of the Bureau of Weights and Measures, in accordance with the following motion prepared by Dr. Ward:

Resolved, That the bar prepared by the United States Bureau of Weights and Measures as a standard for micrometry be accepted, and that a sub-committee of three be appointed to secure copies on glass for such societies as may desire them.

Resolved, That the following report be tendered to the American Society of Microscopists: The national committee on micrometry having received from the United States Bureau of Weights and Measures an excellently ruled bar designed and tendered by the Bureau as a standard for micrometry, and believing that such a standard should be subject to the approval and sanction of the Society, hereby tenders the standard to the Society with the recommendation that it be accepted and adopted as a basis for future studies and discussions in micrometry.

PROF. HILGARD'S REPORT.

CENTIMETER SCALE A, 1882.

This scale is divided into ten millimeters, each division being marked by three lines distant from one another ten microns and the measurement is to be made from the mean position of one triplet of lines to that of another. The first millimeter is again divided in the same manner into tenths of millimeters. The first tenth of a millimeter is subdivided into ten spaces of ten microns each. There are thirteen of these lines at the beginning of the centimeter, the first tenth of a millimeter being measured from the mean of the first three

to the mean of the eleventh, twelfth and thirteenth.

The scale is engraved on a piece of platin-iridium made by Matthey, and containing 20 per cent. of iridium. The coefficient of expansion of this metal from o to 100 degrees C., according to Ste. Clair Deville and Mascart [Annales de l'Ecole Normale, second series, tome VIII., p. 9], is 8.778μ per degree centigrade per meter. But at ordinary temperatures it is no doubt somewhat smaller. The whole centimeter has been found to be 2.18μ shorter than Glass Centimeter No. 1 (Dec. 25, 1878), at 65 degrees F. and also at 70 degrees F., the two centimeters having sensibly the same coefficient of expansion at that temperature. This Glass Centimeter No. 1 (1878, Dec. 25,) is 1.61 microns longer than a mean centimeter of glass, decimeter scale No. 4. This mean centimeter is provisionally taken to be 2.80μ too long at 70 degrees F. We can refer this length to a wave length of light more accurately than to the meter-namely, the correction just given makes the wave length of Kirkhoff, 1,200.6, to be 0.5624918μ in air at 30 inches reduced pressure and 70 degrees F. (This value is subject to future correction).

The corrections to the different divisions of the scale A, 1882, at 70 degrees F. are as follows:

Eleventh line to twelfth line, too long..
Twelfth line to thirteenth line, too short...

..0.08μ

.0.34

..0.05

....0.09

.0.41

0.20

.0.39

.0.19

...0.05

.0.20

.0.18

.0.23

A Study of the Centimeter, Marked "A," Prepared by the U. S. Bureau of Weights and Measures for the Committee on Micrometry.

At its last session, this Society conferred the honor of appointing me as its representative upon the general committee upon micrometry, representing the various microscopical societies of this country. This committee has, through its chairman, President Barnard, of Columbia college, and its Secretary, Dr. Ward, of Troy, obtained from Professor J. E. Hilgard, Superintendent of the United States Coast Survey and Director of the Bureau of Weights and Measures, a standard centimeter ruled upon a platin-iridium surface which appears to satisfy every requirement essential in a standard unit of measurement. This standard was sent to me by Dr. Ward at the beginning of the present year, with the request that I should compare it with the one-hundredth part of the meter in my possession, the relation of which to the Metre des Archives at 62 degrees Fahrenheit has been definitely established. Inasmuch as no special investigation of the coefficient of expansion of this particular plate of platiniridium, attached by silver solder to a plate of brass had been made, I was requested by Dr. Ward to undertake this investigation.

The standard was received by me January 20 of the present year. In its examination with a half-inch objective supplied with a Tolles' opaque illuminator, it became at once apparent that the defining lines are of the most beautiful character. I do not think I have ever succeeded in producing lines upon a metal surface quite equal to the lines upon this plate. The surface of the platin-iridium does not appear to be quite as well prepared as it is possible to prepare a surface of tempered steel, but from some experiments which I have

since made with pure iridium it is evident that this metal, with all its other good qualities, is not well adapted in this respect. This defect is, however, apparent rather than real, since it only affects the definition of the graduations in certain parts. I found a few scratches upon the plate, especially near the third millimeter. It should be noted that it is extremely difficult to polish this metal without leaving traces of the polishing material. While this plate was in my possession, its surface was not touched, even to remove the particles of dust which accumulated upon it, except with a camel's hair brush. The graduated surface is not quite parallel with the lower surface of the brass plate. I found it necessary to cement to one end five thicknesses of tissue-paper.

Before proceeding to give an account of the results obtained in this investigation it will be necessary to allude briefly to the original unit with which this standard has been compared. The original basis of this unit is a meter upon copper prepared for me by Professor Tresca, of the Conservatoire des Arts et Metiers at Paris. This meter was transferred from meter No. 19 of the Conservatory, at 2 o'clock on the morning of Feb. 6, 1880, and its relation to this standard was determined by a sufficient number of comparisons.

According to the report of Professor Tresca, this meter was found to be, by comparison, with the Metre des Archives 118.9μ too long at 13.70° C.

Since it seems desirable that all units of measurement shall be referred to a temperature near the mean temperature at which scientific observations are usually made, I have selected 62° Fahrenheit or 16.67° C., since this is the temperature at which the Imperial Yard is a standard. In order, therefore, to determine the length of the Tresca meter at 16.67° C., it became necessary to determine its coefficient of expansion with great care. It will not be necessary to describe the various unsuccessful attempts which were made to determine this coefficient with precision. It is sufficient to say that I have been unable to obtain satisfactory results by immersing the bar in a liquid. The method which was finally adopted seems to meet every difficulty. The line meter was compared at extreme temperatures with an end-measure meter immersed in melting ice. The details of this investigation will be found in the forthcoming volume of the Proceedings of the American Academy of Arts and Sciences.