88 Mr. J. J. Waterston on certain Thermomolecular 



water-node, and the static series to the ether-node. The obser- 

 vations above 0°, both by Regnault and Faraday, point to the 

 ether-node. The line of sulphide of carbon cuts the zero ordi- 

 nate at 0*712. The points given by the observations on hydro- 

 chloric ether range in a regular curve ; so they cannot strictly be 

 held to conform ; but the hydrobromic and hydriodic ethers give 

 points very exactly in a line. The methyloxalic-ether points 

 range pretty well in a line that is directed to the water-node. 

 The pyroxylic-spirit points are not so exactly in range as the best 

 of the others : two lines were drawn ; the intersection of one is 

 at T18, and of the other at 1*27 ; the mean is 1*225 ; so there is 

 no doubt that its node is the same as that of water. 



§ 21. It will be remarked (Plate I.) that above the boiling- 

 point the observations decline below the straight line; this is quite 

 consistent with the deviation from the law of Mariotte disco- 

 vered by Regnault. The pressure is less than the density would 

 give if that law were maintained at high pressures ; and, vice 

 versa, the actual density is greater. 



§ 22. I have only my own observations (made fifteen years ago) 

 in sealed tubes to present as evidence of the law of density at 

 high temperatures, and in a few cases up to the point of transi- 

 tion. The chart with the points of the original observations 

 laid down will be found in the c Archives ' of the Royal Society 

 for 1852, a copy of which is now before me; audit is gratifying 

 to find that the line passing through the sulphuric-ether points 

 (between 125° and 190°), produced downwards to meet the ordi- 

 nate at —274°, cuts it at 0*79, which, it will be admitted, is very 

 close to the mean from Regnault's observations published in 

 18G2. The tube observations were made, computed, projected 

 on the chart, the line drawn through them, and the copy lodged 

 in the Archives of the Royal Society ten years before the idea of 

 such a nodal point existed. The line is placed, in Plate III., 

 with the letter w across it. Its boiling-point was 41°, and spec, 

 grav. 0*737 at 30° C, so it was not chemically pure. Seven 

 graduated tubes filled in different proportions were simultane- 

 ously observed up to 190° C, this being five more than is ma- 

 thematically required to obtain results for liquid volume and 

 vapour-density from perfect observations, ample means were 

 afforded to check the accuracy of their graduation, and of testing 

 the identity of physical influences to which they were subjected 

 during the observations. Four of them showed complete ac- 

 cordance. 



§ 23. The second trial, with improved heater, was made with 

 alcohol that had 19 per cent, water and boiled at 81°. Six tubes 

 were employed, and the curves of liquid volume were established 

 by repeated trials up to 216°. The accordance of all these 



