180 Dr. Kuenen on the Critical Phenomena of 



from C 2 H 5 I. But the difference may be due to the method 

 of determining the critical state as well. In my experiments 

 the retardation is entirely annihilated by carefully stirring 

 the gas, which may have lowered the critical temperature. 



Results obtained with Nitrous Oxide. 



18. Similar results were obtained with this gas. The 

 increase of pressure, however, was still less. At 20° C. it 

 amounted to 0*15 atm. =11 cm. of mercury. The impurity 

 of the gas may be estimated below 0*0002 : for the critical 

 temperature I obtained 35°'95 and 36 o, 05. The true value 

 will be 36°'l C. probably. 



19. Viliard, who has prepared pure nitrous oxide by a 

 similar process*, fixes the critical temperature of his gas at 

 38°*8 C. But this value was obtained by the indirect method 

 of measuring the densities. The liquid surface disappeared 

 below 38° C, but he does not mention an exact value. This 

 gas must have been very pure, judging from the regular 

 phenomena obtained near the critical point, and the purity of 

 the carbonic acid prepared in the same manner. Viliard 

 states that in a U-tube in which two quantities of liquid 

 N 2 were separated by a column of mercury, a pressure of a 

 few centimetres sufficed to liquefy one of the portions com- 

 pletely. But this test is not so delicate as the one given 

 above, viz. the entire increase of pressure from the beginning 

 to the end of the liquefaction f. 



20. The constancy of my results with the pure substances 

 and the smallness of the admixture left, justify me in accepting 

 an influence of the admixture upon the behaviour of the 

 mixtures of a similar small amount. 



* Viliard, C. R. cxviii. p. 1096 ; Journal de Phys. [3] iii. Oct. 1894. 

 In a letter Dr. Viliard tells me that his estimation of the critical 

 temperature has been too high. In turning- a tube with N 2 at 36 0, 5 O., 

 the liquid surface disappeared and the tube filled itself with a blue mist. 

 Now as stirring comes to the same as repeated turning, it is quite natural 

 that I obtained a value for the critical temperature below 36°'5 O. I 

 have not noticed the blue mist more than o, 2 or o, 3 above 36 o, O. 

 At any rate the disagreement between Villard's result and mine has been 

 reduced to a few tenths of a degree. But it appears from Villard's 

 measurements of the densities of N 2 0, that the behaviour of this substance 

 near its critical point is less normal, so to speak, than the behaviour of 

 carbonic acid. One feels inclined to ascribe this to the influence of minute 

 impurities. 



t In filling a tube with a gas, which is compressed to a liquid after- 

 wards, one works under unfavourable conditions for obtaining a pure 

 substance compared to what can be obtained with liquids. The isopentane 

 prepared by Prof. Young (Phil. Mag. [5] xxxviii. 1894, pp. 569-572) 

 is remarkable in respect to purity. Here the increase of pressure 

 appears to have disappeared entirely. Experiments by Battelli and others 

 prove that even a pure liquid is not easily obtained. 



