bar was cut off and the other two-thirds when forced and well 

 annealed had the following dimension, : Length. 95 cms.: l.readth, 

 2-5 cms.; thickness, 2 cms. Its weigh' i.ht'imln 15 legs ui.i 

 its specific gravity, 21 '048. The bar was next rolled between 

 polished cylinders and brought to very nearly the required dimen- 

 sions, of length 410 cms., breadth 2'1 cms., and thickness -5 cms. 

 Finally, a perfectly rectangular form was given to the bar bv 

 drawing it several times through a hole in a plate of steel. Each 

 time the bar was heated to the highest temperature by a very 

 large gas blowpipe fed with oxygen, and it was passed through 

 the same hole until a force was no longer required to draw it 

 nd it no longer underwent a sensible elongation. Each 

 time when hammered, rolled or drawn, the metal, heated 

 either in a great muffle or by the blowpipe, was cleaned by melted 

 borax and by concentrated chlorhvdric acid. The rule ought to 

 receive from the makers, MA!, ibainn. r, the high degree of polish 

 of which it is susceptible, and which h, its present rough eondi- 



M. Devil le in commenting on this communication gave the fol- 

 lowing result of an analysis of a portion ,,f this alloy : Platinum, 

 89-42; iridium, 10-22; rhodium, -16; ruthenium, -10; iron, -06; 

 total, 99-96. The density computed from that of these compo- 

 nents was 21-515; that observed, 21-516. A second analysis 

 gave by computation 21-51; by observation, 21-508. He then 

 two tubes of the same alloy made from the portion cut 

 off from the bar as described above. These tubes are closed by 

 sph ical cups, of whi one term ites in a pillary tub. vhich 



manometer. Each of these tubes is more than a meter long and one 

 of t hem has a capacity of over a liter. This is to serve as an air ther- 

 mometer for the determination of the boiling points of various 

 liquids under known pressures. These tubes will have on their sur- 

 faces two lines a meter apart. The deviation compared with that 

 of the second tube which will be kept in ice, will enable the exact 

 • to be calculated. This will carry out a Mig option 

 long since made by Regnault, and whi.-h lias been applied to the 

 measurement of high temperatures. It will be interesting to 

 the measurement of the lowest temperatures 

 so far attained. These researches have an important bearing on 

 metrologr. All -ubs; :l r;ce- capable of bei,c tempered, as glass 

 and steel undergo a change with time. Others which like zinc 

 with changes of temperature also change with time. 

 Hyaline quartz, according to Mr. ' ,,! from this 



change. I> ', steel, zinc or quartz? 



This question will be solved by weighing the quantity of water 

 - will contain tit o°. After all the experi- 

 tnents in ..;,.,! between very wide 



limits tie capacity of the tube will be again measured. Each 

 - : . : ' .■ ■• ; ' _ ■:..■■• -•- 



sent the volume of the tube corresponding to a change of one- 



