90 Mr. J. J. Waterston on certain Thermomolecular 



cuts the axis at —274°. The dichloride was also tried, in two 

 tubes but the points did not range well; the general trend was 

 parallel to the chloride. It may be that the true line converges 

 to —274°, certainly not to the ether-node. 



Vapour- density and Liquid volume at Point of Transition. 



§ 29. The following are the extracts from an unpublished 

 paper giving an account of these observations, made in 1852. 



" The chief object of the following experiments was to as- 

 certain if the law of density of saturated vapours holds good up 

 to that point where, according to M. Cagniard de la Tours's in- 

 teresting researches, the liquid condition seems to terminate 

 suddenly. The observations were made on the same principle 

 as those which were the means of detecting the general law of 

 density, the details of which have been communicated to the 

 Royal Society. The tubes were mostly from 2 to 3 inches in 

 length. The graphical method was employed to compute vapour- 

 density and liquid-volume (see Phil. Mag. vol. xxvi. Appendix II. 

 to paper on Expansion of Water) . 



" The following was the method employed to heat the tubes 

 and take the temperatures. A glass funnel, about 3 feet long 

 and 1 inch in diameter, was fixed vertically over a large Argand 

 oil-lamp. One of the prepared tubes was placed in a small brass 

 wire frame arranged and fitted so as to slip into the top of the 

 funnel and hold the tube fixed steadily at the centre of the cur- 

 rent of heated air about 4 or 5 inches below the top. By in- 

 creasing the flame of the lamp gradually, the transition tempe- 

 rature was attained, and by careful adjustment was kept nearly 

 steady. The liquid volume being noted, the tube was taken out 

 and a thermometer put exactly in its place. The mercury, quickly 

 rising, became steady, the temperature was noted, and thermo- 

 meter removed. A second tube, having a different proportion 

 of its volume filled with the same liquid, was slipped into the 

 same place ; then a third, and a fourth. 



" The state of the liquid in the tube was closely examined through 

 a watchmaker's lens by transmitted light. The tubes were ge- 

 nerally of hard glass \ inch bore and ^ inch thick. When 

 carefully sealed they did not burst at the transition -point of any 

 of the liquids, except water, which burst the tube at about 330°, 

 first corroding the glass in a ring at the top of the liquid. 



"The liquids observed by this method were ether, alcohol, sul- 

 phide of carbon, water, chloroform, glacial acetic acid, dichloride 

 of sulphur, turpentine, and sulphuric acid. 



"Ether. — Boiling-point 41°; four tubes; transition-point 221°; 

 density of vapour 0*19 to 0'23 ; density of liquid 043 by third 

 and fourth tubes (unity at 29 c *5). By two of the tubes the den- 



