134 EEPOET— 1886. 



The nitrogen witli wMcli the tube had been filled before the experiment 

 ■was expelled by CO2 till the CO2 was entirely absorbed by potash ; this 

 gives the amount of nitrogen left at the highest tempei'ature of the experi- 

 ment. After the furnace had quite cooled the tube was again filled with 

 nitrogen at the temperature of the room, and the amount of nitrogen at 

 this temperature determined in the same way. The expansion of the 

 nitrogen is thus known between the two temperatures ; the highest 

 temperature is thus easily calculated on the faith of the accui'acy of Gay- 

 Lussac's law for nitrogen through this range of temperature. As an 

 example of this method they apply it to the case of mercury vapour, and 

 find in two experiments 6'89, 6'76, as against the calculated number 6'91. 



The method described above for determining the temperature may be 

 called the nitrogen-thermometer method ; its applicability has been amply 

 justified by further comparisons of its densities with those of other gases 

 at still higher temperatures, with, among these, the gases hydrogen, 

 oxygen, mercury. 



The Behaviour of Iodine at High Temperature. 



In the meantime (year 1879) Crafts, using a modification of V. Meyer's 

 apparatus, found for chlorine no alteration of density, or at the most only 

 a few hundredths at the highest temperature of the furnace ; but for bro- 

 mine, which for Brg should have density 5"7, was found 4"39 at the 

 highest temperature ; while the density of iodine, which for I2 should be 

 8"795, was found reduced to 5'93.^ Thus Crafts found vapour- density of 

 iodine reduced in ratio 1'5 to 1 ; of bromine in ratio 1 2 to 1 ; and of 

 chlorine very slightly reduced, if reduced at all. 



V. Meyer also found the vapour-density of iodine reduced in ratio 1'5 

 to 1, being abnormal above 590°. 



Crafts and Meier,^ by a quite different experimental method from that 

 used by V. Meyer, arrived at results which showed that the temperatures 

 were inaccurate in Meyer's experiments with iodine, and that whereas 

 according to V. Meyer's figures the vapour-density of iodine remains 

 constant between 1000° and 1570°, Crafts and Meier ^ show that it con- 

 tinually diminishes as the temperature rises up to 1400°, when it has a 

 density less than two-thirds the density required by the formula I2. Since 

 then Crafts and Meier ■* extended their experiments to higher tempera- 

 tures, operating under reduced pressure. They find that they get a 

 vapour-pressure of iodine above 1300° (at -1 atmo pressure), which is 

 near to half that for I2 — namely, about 4'6 — and which remains neaiiy 

 constant, slightly diminishing for all temperatures up to 1400°, the 

 curves showing the vapour-densities as ordinates and the temperatures 

 as abscissae. 



Deville and Troost,^ on referring back to an experiment made many 

 years ago (in 1860) on the density of vapour of selenium by comparison 

 with that of iodine at some very high temperature, find a note appended, 

 to the effect that there must have been some mistake made in tlie weight 

 of iodine remaining in the flask, for with the number given, Oil gram, a 

 temperature of nearly 2000° would be attained ; they in this communica- 

 tion recognise that the experiment was accurate and that the smallness 

 of the weight of iodine was due to the abnormal diminution in the vapour- 



' C.B. xc. p. 183. - Hid. p. 690. ' Hid. 



* Ibid. xcii. 39. * Ihid. xci. pp. 54, 83. 



