140 



SCIENCE 



[N. S. Vol. XXXIV. No. 



metallic atoms, could easily lose electrons. 

 The mass would then contain an abundance 

 of mobile or free electrons and in such case 

 possess high electrical and thermal con- 

 ductivity, metallic luster, etc. 



Turning now from theory to facts, the 

 case of ammonium amalgam demands con- 

 sideration at, once on account of its his- 

 torical importance. This remarkable sub- 

 stance was discovered practically simul- 

 taneously and independently by Seebeck, 

 and by Berzelius in 1808 ; curiously enough, 

 in just the same year that Davy isolated 

 sodium and potassium from their hydrox- 

 ides. Two years later Davy, in 1810, com- 

 pared ammonium amalgam with the amal- 

 gams of sodium and potassium and was led 

 to announce his famous ammonium hy- 

 pothesis; the radical ammonium was anal- 

 ogous to the alkali metals and was said to 

 exist in metallic form, united with the 

 mercury, in ammonium amalgam. Ber- 

 zelius and Ampere also supported this 

 view. Some years later, after the discov- 

 ery of other radicals, Dumas and Liebig in 

 a joint paper gave Davy's idea a much 

 more general form. They wrote : ' ' Organic 

 chemistry possesses its own elements which 

 sometimes play the part of chlorine or 

 oxygen, sometimes, however, also, that of a 

 metal. Cyan, amid, benzoyl, the radicals 

 of ammonia, the fats, the alcohols and their 

 derivatives, form the true elements of 

 organic nature." But the hypothesis of 

 the metallic nature of ammonium in the 

 amalgam did not pass unchallenged. Gay- 

 Lussac and Thenard concluded that the 

 so-called amalgam is only a mixture of 

 ammonia, hydrogen and mercury; a 

 view subsequently shared by many others, 

 among them Seely, who found the volume 

 of the inflated mass to be inversely propor- 

 tional to the pressure upon it. The case 

 against the metallic ammonium hypothesis 

 was made still stronger by the evidence 



furnished by an experiment by Landolt in 

 1868. If the amalgam is really analogous 

 to sodium amalgam, if the radical actually 

 has the properties of a metal, it should 

 readily precipitate from solutions of their 

 salts metals of smaller solution tension; 

 but, in the test, Landolt could precipitate 

 neither copper nor silver with ammonium 

 amalgam. 



The first really convincing evidence in 

 favor of the ammonium hypothesis was fur- 

 nished by LeBlanc in 1890. LeBlanc elec- 

 trolyzed a solution of an ammonium salt 

 with a mercury cathode. The apparatus 

 was so arranged that simultaneous meas- 

 urements of the polarization potential 

 could also be made. This potential rose in 

 a few minutes to a maximum which was 

 nearly as great as that given by a sodium 

 salt. The really important result, however, 

 was observed after the polarizing current 

 was cut off. The mercury cathode, which 

 showed the inflation characteristic of am- 

 monium amalgam, was still strongly elec- 

 tro-negative toward the solution and re- 

 mained so for from ten to twenty minutes. 

 That this effect was not due to hydrogen 

 was shown by the fact that the hydrogen 

 polarization potential was considerably 

 smaller and that it fell off almost as soon 

 as the current was interrupted. These ex- 

 periments of LeBlanc, based as they were 

 on the sound principles of electro-chem- 

 istry, gave a new impetus to the ammonium 

 hypothesis. Coehn, in 1900, reasoned that 

 if ammonium amalgam gave the high po- 

 tential found by LeBlanc, it surely ought 

 to precipitate copper and silver, and that 

 Landolt 's experiments should succeed. 

 But Coehn failed exactly as did Landolt! 

 Coehn next found that at very low tem- 

 peratures, or even at zero, the amalgam 

 was much more stable than at room tem- 

 perature, and would precipitate copper 

 from copper sulphate without difficulty. 



