OF ARTS AND SCIENCES. 281 



made by Brunner Freres, of Paris, which is presumably a copy of the 

 line-meter derived from the meter of the Observatory. The meter by 

 Brunner Freres, the tracings on which are very good, is taken as the 

 present provisional standard. 



The buildings of the International Bureau are situated on the 

 summit of a high hill at Breteuil, on the direct road from Paris to 

 Versailles. The location is an admirable one in every point of view. 

 During my recent visit to Paris, a very kind letter of introduction 

 from Professor St. Claire Deville to Dr. Pernet, the Director of the 

 Bureau, led to a very interesting visit to this establishment. There 

 are three principal observing-rooms, perhaps eighty meters square. 

 The walls are composed of corrugated metal, the waves running lon- 

 gitudinally. The rooms are lighted by a circular skylight only. 



In one room is found the apparatus for comparing standards of 

 weight, under the charge of Dr. Marek. There is here a marvel- 

 lously perfect apparatus for weighing in a vacuum, made hy Bunge, of 

 Hamburg. The observer is enabled to perform every part of the 

 operation of weighing, stationed at a distance of six or eight meters. 

 There is also a fine balance by Rupprecht, of Vienna. 



In a second room, the instruments for determining the coefficients 

 of expansion are mounted. This department is under the charge of 

 Dr. Benoit. Both the method by immersion in a liquid, and the method 

 of M. Fizeau, are employed. The latter method is described in the 

 Comptes Rendus for I860. It is also described in the Proceedings of 

 the Royal Society of Nov. 30, 1866. I quote from the Proceedings: 

 <l Iu M. Fizeau's observations, he has availed himself of the possibility 

 of forming Newton's rings with the monochromatic sodium light, when 

 one of the interfering rays is 52.205 waves in advance of the other, a 

 fact which, conjointly with M. Foucault, he announced in 1849. Using 

 the length of a wave of sodium light 0.0005888 mm. [0.000023 18' 1 "-] 

 as a standard of measure, the position of a ring being observable to within 

 •j^ of the distance between two consecutive rings, the variation of the 

 distances between the two surfaces producing Newton's rings can be 

 measured within 0.U00<>2944 mm. or 0.0001159 in - " The results by this 

 method, and those obtained by immersion in a liquid, are found to show 

 a good agreement. The chief objection to its use is found in the fact 

 that the method is only applicable to pieces of metal not much exceeding 

 one centimeter in length. It must be assumed that the whole bar has the 

 same coefficient as the small portion of it upon which the observation 

 rests. In the discussion of the coefficient of the Metre des Archives, it 

 will be interesting to compare the value finally derived with the value 



