208 Transactions of the Boyal Society of South Africa. 
feiTocyanide). K^Fcy, after KMnO^ not in excess, gives an amorphous 
dull-grey compound, so that Barzilowsky's blue substance is a ferri- 
cyanide. On treatment with ammonia internal reduction occurs and 
a little benzidine and ferrocyanide appear in the aqueous extract, and 
the crude diimine base is left insoluble. The blue compound gives an 
orange colour with H2SO4 (the base, crimson as usual) and is apparently 
SC.^H.oN, + H3FeCy6. Found Fe - 8-6 per cent, instead of 9-7 per 
cent. When reduced with KHSO., benzidine sulphate was the only 
product. 
The base from Barzilowsky's compound was boiled out with water, 
dissolved in much alcohol, reprecipitated, and again dried at 100°, and 
was nearly free from ash. 0-1232 gave 0-3470 CO^ and 0-0690 H^O ; also 
0-0020 Fe,03. 0-239 used 11-3 cc. N/5 HCl. Allowing 0-0022 for HFeO,, 
this gives C = 78-3, H = 6-3, N - 13-5. The substance melted at 255-260, 
and behaved exactly as Willstatter's diimine does. 
The following experiments show that despite the poor analytical 
figures for the dried substances, the blue compounds are really salts 
of diphenoquinone-diimine : — 
(1) Quantitative Beduction : 1 gram benzidine in HCl at 40° (in 
about 1 litre water) was treated with 2*5 grams bichromate and the 
blue paste digested for 20 minutes (when all benzidine is oxidised). 
Stannous chloride was added, and when the mixture became clear the 
tin was removed by H2S. The tin precipitate was extracted with soda 
and alcohol to recover adherent benzidine and the extract added to the 
rest. Sodium sulphate was added in excess and the benzidine-sulphate 
allowed to crystallise out during two days. 1'334 gram was obtained. 
The mother liquor was taken to small bulk and acetylated, giving 
0-0825 gram of white diacetylbenzidine of mp. 310° (uncorr.). The 
acetylation liquor still contained a trace of benzidine recognisable with 
CrOg. Nevertheless, 93 per cent, of the benzidine taken was recovered 
as sulphate and acetylderivative, thus leaving practically no possibility 
of there being another reduction-product. Similarly, when the diimine 
hydrochloride is reduced by SO2 the only product isolated was benzidine 
sulphate. On the other hand, when the dried blue substance is reduced 
only a comparatively small quantity of benzidine is obtained and most 
of the substance remains insoluble. 
(2) Isolation of Diimine. — In order to avoid decomposing the diimine, 
the blue paste obtained as described above, or Barzilowsky's compound, 
was decomposed in situ without filtration with excess of dilute caustic 
soda, and filtered off, using the pump, and thoroughly washed out with 
cold water. (Owing to the slimy character of the product this operation 
takes many hours.) It was then dried for a week over H2SO4 in vacuo 
and the product pulverised and extracted with a mixture of ether and CSg 
