76 Scientific Intelligence. 



its cobalt and is converted into cobalti-nitroso-/?-naphthol, 

 [C 10 H c O(XO)] 3 Co, of a purple-red color. The same body is ob- 

 tained when a neutral or acid solution of a cobalt salt is added to 

 a solution of nitroso-/i-naphthol in alcohol or acetic acid. It is 

 remarkably stable, resisting acids, alkalies, oxidation and reduc- 

 tion reagents, etc., to a marked degree. It dissolves in fuming 

 nitric acid and in strong sulphuric acid, but is precipitated un- 

 changed on dilution. Boiled with strong alkalies, it is only slow- 

 ly attacked. Hot acetic acid (50 per cent) dissolves it sparingly, 

 but deposits it again completely on cooling. It is slightly soluble 

 in strong alcohol, more so in dilute alcohol and readily in aniline 

 and phenol. Nickel salts give under similar circumstances nickel- 

 nitroso-/?-naphthol as a brown-yellow precipitate difficultly soluble 

 in water and alcohol and from which hydrochloric and sulphuric 

 acids easily remove the nickel leaving the nitroso-/3-naphthol, 

 which in presence of sufficient acetic acid goes into solution. As 

 therefore a nickel salt produces no precipitate in an acetic solution 

 of nitroso-/5-naphthol containing hydrogen chloride, the authors 

 recommend the following method of separation : To the solution 

 containing the nickel and cobalt as sulphates or chlorides, acidu- 

 lated witli hydrochloric acid and slightly warmed, a hot solution 

 of nitroso-/?-naphthol in 50 per cent acetic acid is added. The 

 precipitate is allowed to deposit and the supernatant liquid tested 

 with more of the reagent. After some hours, the precipitate is 

 filtered off, washed first with cold, then with hot water, and then 

 with a 12 per cent solution of hydrochloric acid to remove the 

 whole of the nickel. After drying, a few grains of oxalic acid free 

 from ash are added, and the precipitate is burned as usual in a 

 tared crucible. The residue is ignited in a current of hydrogen 

 gas and weighed as metallic cobalt. The nickel may be deter- 

 mined in the filtrate as usual. By determining both metals in an 

 aliquot part of the solution by precipitation with potassium 

 hydrate and reduction to the metallic state, and then estimating 

 the cobalt in another similar portion as above, the nickel is readily 

 determined by difference. In 20 c.c. of a solution containing 

 0-0382 gram cobalt, 0-0380, 0-0378, and 0-0379 gram were obtained 

 in three experiments. Fifty c.c. of this cobalt solution, mixed 

 with 100 c.c. of a nickel solution containing 0-2097 in., gave 

 0*0947 gram Co instead of 0*0955. As to delicacy, 0-2 c.c. of a 

 cobalt solution containing 0-00004 gram Co, mixed with 5 c.c. of 

 a nickel solution containing 0-0105 Ni, gave when mixed with the 

 solution of nitroso-/3-naphthol, an immediate turbidity, and after 

 a time flocks of the cobalti-compound separated. — Ber. Berl. 

 Chem. G-es., xviii, 699, March, 1885. G. f. b. 



5. On a ready method of preparing Tartronic acid. — Although 

 many processes have been devised for preparing tartronic (oxy- 

 malonic) acid, yet owing to the cost of the crude material or the 

 very small yield, this acid has been but little studied. Pixistee 

 has recently observed the formation of this acid by the action of 

 sodium hydrate upon ethyl trichlorlactate according to the equa- 



