TRANSACTIONS OF SECTION B. 307 



Briihl proposed working the Hofniaiin method at a very low pressure by 

 ■employing a tube 1*6 metres long with only a small quantity of substance, and 

 was therefore able to make determinations at temperatures far below its boilin* 

 point. He also made the following suggestions : — 



1. In order to eliminate the troublesome element of the tension of mercury- 

 vapour (without using two tubes, as Grabowski did), heat the column to the 

 required temperature, note its height, then allow it to cool ; introduce the substance, 

 and heat again to the same temperature till the height is constant. To ensure 

 uniformity of level in the bath keep it full to overflowing. 



2. Before the first reading of the mercurial column, a small piece of thin 

 glass is passed up to liberate any air that maybe contained in the mercury. 



3. To make a mark on the tube a little above the vacuum mercury level, and 

 then only to calibrate about 150 m.m. down from that point ; then, to find the 

 total volume, add the variable volume below the mark to the fixed volume above 

 the mark. 



Muir and Suguira, in 1877, finding that sometimes the weight of the inner tube 

 caused the groove in the india-rubber disc to so far close as to prevent the escape 

 of the mercury while heating the substance, used a plain india-rubber disc, which 

 was fastened to the bottom of the trough, a disc of cork intervening. Oom- 

 munication between the mercury in the tube and that in the trough was maintained 

 by means of a short piece of glass tubing bent at right angles. A second tube, 

 long enough to stand slightly above the level of the mercury iu the trough, served 

 to carry off from the space between the two tubes the condensed liquid and 

 excess of vapour. They adopted Hofmann's original method of passing the steam 

 in at the top of the outer tube, but used a small tube passing through a perforated 

 cork in preference to one fused to the end. 



Briihl has this year proved, by most carefully conducted experiments, that the 

 Hofmann method cannot be used above 220°, owing to the great and rapidly 

 increasing vapour-tension of mercury; but has omitted the grave objection to 

 his own method. Playfair and Wanklyn called attention in 1861 to the fact that 

 Bineau, in 1846, pointed out that in vapour-density methods, at very reduced 

 pressures, slight errors in the readings of the mercurial level introduce very serious 

 errors into the result. 



In the Overflow methods, which are in reality modifications of the Gay-Lussac, 

 seeing that they are performed with known weights of substance, the first name 

 is Hofmann, who in 1860 gave a very meagre description of his apparatus, when 

 he wrote that he used an U tube heated in a paraffin bath, and estimated the 

 volume of the vapour by the mercury expelled. Werthein, in 1862-64, in his 

 papers on Ooniin, gave full details of his method, in which he used two tubes 

 suspended side by side in a flask. 



Watts, in 1867, employed a globe with a ground neck, into which an outlet 

 tube, reaching nearly to the bottom of the globe was accurately fitted. The globe 

 being filled with mercury, and the substance introduced, the quantity of mercury 

 expelled on heating served as a basis for calculating the volume occupied by the 

 vapour. Victor Meyer, in 1876, introduced two very important alterations, he 

 avoided the vapour-tension of mercury by using fusible metal, and placed the 

 outlet at the bottom of the bidb. His experiments at that time were all made 

 in the vapour of boiling sulphur, but Graebe, last year, wishing to employ a higher 

 temperature, used phosphorus pentasulphide, which boils at 530°. 



Frerichs, in 1876, used mercury in an apparatus similar in principle to that of 

 Watts, but employed an inverted flask, and brought the exit-tube, which was 

 furnished with an inverted siphon, through a suitable outlet in the bottom of the 

 bath. 



Goldschmiedt and Oiamician, in 1877, used mercury with the simpler bulb of 

 Victor Meyer, but added a small side-tube to the outlet, so that the mercury 

 expelled could be weighed from time to time during the heating. Victor Meyer, 

 in the same year, modified the shape of the bulb, but heated it in a tube similar to 

 that employed by Greville Williams in the Gay-Lussac determinations, but of 

 sufficient length for the upper part of the tube to serve as condenser. 



x2 



