Foreign Metals on the Electric Conducting Power of Mercury. 35 



first and then filled the tube with them, as Dr. Siemens did with 

 his zinc amalgams (especially with those which partially solidify 

 on cooling), no concordant results could be obtained; for in ma- 

 nipulating in this manner there is a great probability of the alloy 

 in the tube not having the same composition as that in the cups. 

 Now, if we wished, as Dr. Siemens did, to analyse the contents of 

 the tube after having tested its resistance, it would be necessary 

 to remove the connectors which dip into the cups, and then the 

 cups, in order to empty the tube. But in each of these opera- 

 tions the level of the amalgam in the cups would be disturbed, 

 and so cause, in the case of amalgams which partially solidify 

 on cooling, a flowing of a poorer amalgam into the tube, which on 

 being analysed would contain a less quantity of foreign metal 

 than was contained in the amalgam tested, and in this manner 

 the values deduced for the conducting powers of the metals 

 would be too high. That this really does take place we have 

 often found; for in testing the resistance of a partially solidified 

 amalgam, if one of the connectors be removed and then replaced 

 in the tube, the resistance will have materially altered; but if 

 now both connectors be removed, and the tube heated over a 

 ♦lamp so that the amalgam becomes perfectly fluid, and then 

 retested with the precautions we are about to detail, the ori- 

 ginal resistance will again be found. We do not find in Dr. 

 Siemens's paper any description of the method he employed to 

 prevent the error just mentioned, nor any indication of the 

 method he used to analyse his amalgams containing very small 

 percentages of foreign metal. In our experiments* the tube 

 was filled with the requisite quantity of mercury, and the other 

 metal added, the tube then carefully heated over a lamp, at the 

 same time allowing the mercury to flow continually from the 

 one arm to the other. When the metal was quite dissolved, the 

 tube was laid in the trough whilst hot, and the copper con- 

 nectors placed in the wide tubes (these being previously heated 

 when experimenting with amalgams which partially solidify on 

 cooling). 



From the form of the tubes, as shown in our paper, page 173, 

 it is evident that, as the thick connectors reached to the bottom 

 of the wide tubes, they, as it were, closed the ends of the ther- 

 mometer-tubes, and would therefore serve to impede filtra- 

 tion of a poorer amalgam into the tube. The trough being now 

 filled with water, the resistance of the amalgam was tested as soon 

 as it had taken the temperature of the bath. In this manner of 

 experimenting, no filtration of a poorer amalgam into the narrow 

 tubes (which had diameters varying from 1 to 2 millims.) can 

 well take place. It is scarcely probable that, in making the 



* For full details of the particulars taken see our paper, loc. cit. p. 173. 



D 2 



