4 C. Barxis — Experiment with Liquid Carbon Dioxide. 



towards the screen G). The linear focus corresponding to the 

 gas is however, virtual, for it is due to the refraction of a 

 cylindrical shell of glass essentially of wide aperture, and lay 

 about S'^"' behind the tube (i. e. away from the screen G). 



Unfortunately the only tube available was too full of liquid 

 COg, so that the liquid reached the top and could not 

 be safely heated above 30° (critical temperature 31*1°) ; 

 but I traced the total extinction of the gaseous focus, this 

 being crowded out by the corresponding advance of the liquid 

 focus quite into the top of the spherical end of the tube. 



As soon as I get another tube I hope to push this experiment 

 through to a definite conclusion. As far as I have gone, there 

 seems to be small probability for the occurrence of real con- 

 tinuity between COj gas and CO^ liquid at the critical tem- 

 perature. There is continuity between the liquid and a gas 

 which preserves the sarne molecule^ the same molecular struc- 

 ture as the liquid from which it issues. Doubtless at still higher 

 temperatures the gas with the liquid molecule will break up 

 into the true gas with the gaseous molecule, and the fact 

 should be indicated by the sliding lens method sketched in 

 the above experiment. 



Finally it is desirable to specify a superior limit for the 

 molecular weight of the liquid in question. The critical vol- 

 ume of carbon dioxide referred to the normal gas (0° C, 76'^''') 

 is -0043 (Cailletet verified by Sarrau). Thus the, vapor density 

 is 4*3 and the molecular weight by Avogadro's law, 123. This 

 is singularly near CgOg, whose molecular weight is 132. The 

 question thus arises whether Qfi^ gas above the critical tem- 

 perature may not be as justly treated as a gas, as any other 

 vapor under the ordinary conditions of vapor density measure- 

 ment : for one may argue that the intermolecular forces which 

 cause divergence from Boyle's law, etc., in CO^ gas at the criti- 

 cal point, have in Cfi^ gas become intramolecvZar, giving the 

 new molecule greater freedom of motion. In such a case Ofi^ 

 would be the molecule of liquid carbon dioxide. 



rapid march toward infinity. As in the above CO2 tube, the Hquid focus lay about 

 3^'*' ahead of the tube, the smaller refraction of this liquid is manifest, but so far 

 as can be detected by the lens method, the gas focus does not appreciably differ 

 in position from the gas focus for atmospheric air. 



Brown University, Providence, R. I. 



