561 
of Edinburgh. Session 1881-82. 
The numbers agree also very well with the formula 2Bi(CN) 5 
5Bi 2 0 5 , which requires bismuth 79 ‘24, cyanogen 8T8, and oxy- 
gen 12*58 per cent. Various other formulae may be found, agree- 
ing more or less closely with the analytical results, but in each 
the atomic proportion of oxygen to bismuth, in excess of that 
required to form Bi(CU) 3 or Bi(CfST) 5 , must be expressed by a 
formula Bi 2 O a; where x is not less than 5 ; in other words the action 
of potassium cyanide has been an oxidising action, while at the same 
time producing a cyanogen compound. 
4. Unless a considerable excess of potassium cyanide is added, 
but little of the new salt is produced. If the liquid be so nearly 
neutral before adding cyanide that one drop of moderately strong 
potash solution produces a permanent precipitate, the precipitate 
produced by the cyanide consists of nearly pure Bi 2 0 3 , with small 
quantities of the new salt. If there be rather more free nitric acid 
than this, the precipitate produced by the cyanide is rich in an oxide 
of bismuth to be described in the sequel, viz., Bi 4 0 7 (see par. 5). 
In one instance one part of potassium cyanide was added to one 
part of bismuth nitrate, dissolved in the minimum quantity of dilute 
nitric acid, a portion of the liquid was boiled for three or four minutes, 
and another portion for twenty-five minutes. The precipitate pro- 
duced in the latter liquid was darker in colour than that in the 
former, and contained rather less bismuth; the lighter coloured salt 
contained 81 ‘59 per cent, bismuth, the darker salt gave bismuth 
= 80 '89 per cent. 
When from 2 to 2*5 parts potassium cyanide are used for each 
part bismuth nitrate in solution, and the liquid is boiled, the precipi- 
tate very quickly becomes of a uniform puce-red colour, while 
the supernatant liquid is reddish-brown. The best results are, I think, 
obtained with this proportion between cyanide and bismuth nitrate. 
When the cyanide solution is added to a cold solution of bismuth 
nitrate, an almost colourless salt is thrown down, which becomes 
puce-red on boiling; this colourless salt is perfectly free from 
cyanogen, and appears to be bismuthous oxide. If, however, 
anhydrous bismuthous oxide be boiled with potassium cyanide 
solution,* no change occurs, nor does addition of potash induce any 
* This has been confirmed by repeated trials, and by analysis. 
