356 RECORD OF SCIENCE FOR 1886. 



and ammouia seem to follow Dewar's law, that the ratio between the 

 absolute critical temperature and the pressure is constant. The other 

 substances deviate from it, giving ratios which increase with molecular 

 complexity. (J. Phys, February, 1886, II, v, 58.) 



Subsequently these authors extended their observations to include 

 l^ropyl chloride, the three ethylamines, and the two lower normal pro- 

 pylamines, all liquid at the ordinary temperature. Their results con- 

 firmed those above mentioned, that the absolute critical temperature 

 increased with molecular complexity somewhat more rapidly than the 

 critical pressure. Moreover the critical temperatures and pressures of 

 isomeric substances are far from being equal. (C. R., August, 1886, 

 cm, 379.) 



In a lecture at the Royal Institution, Frederick Siemens discussed 

 the phenomena of dissociation with especial reference to questions con- 

 nected with practical heat operation. (Nature, May, 1886, xxxiv, 64.) 



Wroblewski has determined the density of liquified atmospheric air 

 as well as that of its component gases, by measuring the volume of gas 

 given by a known volume of the liquid. At the critical temperature— 

 118°, the density of liquid oxygen is 0.6 ; while at— 200°, at a pressure 

 of only 2 centimeters, the density is 1.24. Hence its atomic volume is 

 less than 14. The density of liquid nitrogen at— 146.6o,the pressure of 

 its saturated vapor being 32i atmospheres, was found to be 0.4552 ; while 

 at— 793°, under a pressure of one atmosphere, the density was 0.83 ; 

 and at— 202O, with a pressure of 0.105 atmospheres it was 0.866. The 

 atomic volume is then about 15.5. Its expansion coefficient is 0.0311 

 at— 153.70, 0.007536 at — 193^, and 0.004619 at— 202°. Atmospheric air 

 on compression behaves like a mixture whose components followed 

 different laws of liquefaction; and when liquified its composition changes 

 with change of temperature or i^ressure. The value of the density found 

 by experiment at — 146.6° and under a pressure of '45 atmospheres was 

 0.59. Calculation gives 0.60 as the value obtained from liquid oxygen 

 and nitrogen. (0. R., May, 1886, cii, 1010.) 



Nilson and Peterssen have described a new method for determining 

 the vapor density of volatile substances and at the same time the tem- 

 perature of the experiment. Four determinations with this apparatus, 

 of the density of the vapor of glucinum chloride at temperatures vary- 

 ing from 1080° to 1502°, gave 2.77 as the mean value. (Ann. Chim. 

 Phys., December, 1886, VI, ix, 554.) 



Nodon has utilized the hygroscopic properties of gelatine in the con- 

 struction of a recording hygrometer. A layer of this substance is fast- 

 ened to the outside of a helix of Bristol board, the inside being pro- 

 tected by a non-hygroscopic varnish. An increase in the atmospheric 

 moisture expands the gelatine and unrolls the helix, as a change of tem- 

 perature does the helix of Breguet's thermometer. The result appears 

 within ordinary limits to be independent of temperature. The hygrom- 

 eter consists of four suQh b^Uces group^(i \u pairs upon the same 



