430 
By means of bringing an excess of pure bro- 
Mine in contact with pure mercury at approxi- 
mately 300°, mercuric bromide was sublimed into 
a convenient receiver, which in turn was heated 
almost to the point of sublimation of the material 
in a stream of nitrogen. On cooling, the nitrogen 
was displaced by dry air. 
From this material, mercuric oxide was pre- 
cipitated by the use of a slight excess over the 
calculated amount of sodium hydroxide. This 
action was brought about in a flask with well- 
guarded openings so that the subsequent reduction 
with hydrogen peroxide or with hydrazine gave 
rise to no loss of the solution in the form of spray. 
The solution being separated from the free mer- 
cury, the bromine was determined as silver bro- 
mide in the usual way. The value 200.63 was 
obtained as the result of eleven concordant deter- 
minations of the ratio HgBr.:2AgBr, no deter- 
minations, of course, being rejected. 
J.I.D. HINDS: Precipitation of the Copper-Arsenic 
Group and the Separation of its Divisions. 
A definite acid concentration is secured. Arsenic 
is precipitated in 2 normal hydrochloric acid 
solution, the other metals in half normal hydro- 
chlorie acid solution. Tin is precipitated as stan- 
nice sulfid. The sulfids of arsenic, antimony and 
tin are dissolved in colorless ammonium sulfid, or 
in ammonium hydroxid and hydrogen sulfid. 
To a portion of the solution add one ninth its 
volume of hydrochloric acid (making it normal in 
HCl since the laboratory acid is about 10 normal) 
and a few drops of nitric acid to oxidize stan- 
nous to stannic ion. Boil the mixture a little more 
than half away in an Erlenmeyer flask, making the 
residual liquid 2 normal, pass a rapid stream of 
hydrogen sulfid until precipitation is complete (5 
to 10 minutes), add enough water to make the 
volume twice the original, making the solution one 
half normal in hydrochloric acid and continue to 
pass hydrogen sulfid until precipitation is com- 
plete (10 to 15 minutes). Filter and wash. 
Transfer the precipitate to a beaker, cover it 
with ammonium hydroxid, pass hydrogen sulfid 
rapidly for a minute, warm, shake, filter and wash. 
The filtrate contains the thioanions of arsenic, 
antimony and tin; the residue contains the sulfids 
of the other metals of the group. 
Treat filtrate and residue in the usual way. 
Time required for the entire process 30 to 45 
minutes. 
WiiuiaM D. HARKINS: The Intermediate Ion Hy- 
pothesis. 
SCIENCE 
[N.S. Vou. XXXV. No. 898 
The values now used for the degrees of dissocia- 
tion of unibivalent salts, such as K.SO,, BaCl, or 
Cu(NO;)., do not represent the degrees of disso- 
ciation at all, but are only the values of the con- 
ductance ratio, if intermediate ions are present in 
the aqueous solutions of salts of this type. In 
recent papers® it has been shown that the solu- 
bility relations of such salts indicate that the 
ionization takes place in two steps as follows: 
1. K,SO,= K* + KS0,. 
2. KSO,;= Kt-+ SO--. 
Potassium sulphate, according to the present 
values used, is 71 per cent. dissociated in its tenth 
normal solution at 25°. Approximate calculations 
made upon the basis of the intermediate ion 
hypothesis indicate that its actual total dissocia- 
tion is about 95 per cent. 
It seems probable that all trvionic salts, acids 
and bases, when dissolved in water, dissociate in 
two steps, and that intermediate ions are present 
in all such solutions. If this is the case, it seems 
self-evident that intermediate ions must be present 
in all aqueous solutions containing salts, acids or 
bases of still higher types. The percentage of the 
salt present as the intermediate ion is zero at zero 
concentration and increases as the concentration 
of the salt in the solution increases. 
The constant k=(K+ X KSO,/K.SO,)  in- 
creases with the total ion concentration and is 
several times larger, in the case of salts, than the 
second constant k= (K+ X SO,;--/KSO,_). 
Solutions of certain salts, such as lead chloride, 
contain an abnormally large percentage of the 
intermediate ion. 
G. R. WHITE and H. Eastwoop: Electrolytic Cor- 
rosion in Ammonium Salts. 
Test pieces of copper, nickel, zinc, tin, iron and 
cadmium were made anodes in solution of am- 
monium chloride, sulphate, nitrate, acetate and 
tartrate containing 75 g. of the salt per liter. 
The anodes were rotated to ensure thorough stir- 
ring. The electrolysis was carried on at room 
temperature for an hour with a current of .35 amp. 
The average current density was 2.8 amp. per 
sq. dm. The results, which are given as percent- 
age efficiency of corrosion, show that corrosion is 
markedly different for different electrolytes and 
that it is affected by changing the current density 
and in some cases by changing the temperature. 
J. W. TURRENTINE and RAYMOND L. Moore: Con- 
tributions to the Electrochemistry of Hydro- 
3 Jour. Am. Chem. Soc., 33, pp. 1807-78. 
