OF ARTS AND SCIENCES. 283 



des Archives," a strip of very thin platinum, terminating in a sharp 

 point, is attached. The reflection of this point from the end of the 

 bar gives the means of observing the point of contact without actually 

 touching thf surface. This method seems to he a necessity in this 

 ca>e, since a statute law forbids contact of any kind whatever. If it 

 was not for this necessary limitation, a much better method could be 

 employed. Still, it is the opinion of M. Tresca that the absolute error 

 of the line prototype can be reduced below 1 fx. = .001 mm. 



On account of the difficulties attending the transfer from an end- 

 measure to a line-measure, M. Tresca has adopted the plan of trans- 

 ferring one line-meter with the utmost precision. This copy, which 

 he calls his " working meter," has occupied his attention for several 

 months. He is confident that ever}' source of error has been elim- 

 inated. This line-meter will be the basis of the final standard, 

 since the Commission will accept without question the definitive trans- 

 fer offered by the French Section after the difficulties with respect 

 to the question of material are removed. On this point there is no 

 dispute. The real issue is this. The International Bureau ask of 

 the French Section one or more bars of pure platinum, and several of 

 platinum-iridium, with the definitive comparison with the " Metre des 

 Archives." The Commission contends that the bars already offered 

 contain two per cent of iron, transferred through the process of < [raw- 

 ing through dies. It contends also, that, by the process of drawing 

 through dies, the bars remain in a state of strain. M. Tresca, on the 

 other hand, contends that it is impossible that these criticisms can 

 hold. 



He admits that iron is transferred during the passage of the bars 

 through the rollers, but this iron is extracted after each of the two 

 hundred passages through the dies. M. Tresca also admits that the 

 bus would be in a state of strain if left in the state in which they come 

 through the dies, but he anneals them finally. As a proof of the cor- 

 rectness of this view, he gets exactly the same coefficient of expansion 

 after each melting and casting of a given bar. Some bars have been 

 melted and recast as many as ten times. 



I heard but little said on the question of the alloy. The standard 

 weights of the Bureau are made of platinum-iridium, and there seems 

 to be no question about them. 



The < iparing rooms of the Conservatory are built as follows. 



First («), there are the thick stone walls of the building; then (b), space 

 filled with hay; (c), wooden walls; (rf), space filled with cotton; 

 (e), wooden walls, covered on the inside with paper. A copper room, 



