PROFESSOR ROSCOE’S RESEARCHES OH VANADIUM. 
325 
was so completely excluded that the alkaline mother liquor from the tetrabasic salt did 
not effervesce on the addition of excess of acid. 
This apparatus consisted of two flasks connected with each other in such a way that 
the liquid contained in one of them could be passed into the other by compressing an 
india-rubber ball, and before the liquid entered the second flask it passed through a wide 
piece of glass-tubing in which a small filter was placed. The apertures were then closed 
by tubes containing solid caustic soda, and the salt could thus be dissolved in alcohol or 
water without any fear of entrance of carbonic acid. 
2-0393 grms. pure V 2 0 5 and 3-7 grms. pure carbonate of soda were mixed and fused 
until no further effervescence occurred. The loss of weight amounted to 1-441 grim, 
whilst the theoretical loss for three molecules of C0 2 is 1-474 grm. The fused mass was 
dissolved in water and boiled for some time in an atmosphere free from carbonic acid, 
and finally precipitated with an excess of strong alcohol. After filtration the clear 
solution containing the caustic soda formed by the decomposition was neutralized by 
standard hydrochloric acid ; it required 4 cub. centims. for saturation (1 cub. centim. 
= 0-0366 grm. II Cl), corresponding to 0-124 grm. Na 2 0. The precipitate was again 
dissolved in a small quantity of water and reprecipitated by alcohol ; this second solution 
needed 4 - 6 cub. centims. of acid for saturation, corresponding to 0-1429 grm. Na 2 0. A. 
third repetition of the process showed that 2 cub. centims. of acid was needed, or 0-062 
grm. Na 2 0. Thus altogether 0-3289 grm. of Na 2 0 was obtained, or nearly half the 
amount required by the formula 2Na 3 V0 4 = Na 4 V 2 0 7 -|-Na 2 O, namely 0*692 grm. 
The alcoholic solutions were free from carbonic acid, and they yielded, on addition of 
silver nitrate, a brown precipitate of silver oxide. They were likewise proved, by evapo- 
ration and treatment with oxalic acid, to be free from any trace of vanadic acid. 
From the above experiment it is seen that the decomposition of the tetrasodium salt 
goes on by degrees, a fresh portion of free soda being found in solution each time the 
salt is dissolved and precipitated ; it is even possible that the precipitation with alcohol 
causes a partial recombination of the caustic soda with the tribasic salt. 
Analyses of Tetrasodium Vanadate. Vanadium determinations : — 
(1) 1*0325 grm. of crystallized pyro-salt was dissolved in a small quantity of water 
and mixed with pure chloride of ammonium. After standing for twelve hours the 
insoluble metavanadate of ammonium* filtered off, washed first with a mixture of a 
saturated aqueous solution of sal-ammoniac and alcohol, and lastly with pure alcohol. 
On ignition 0-2995 grm. V 2 0 5 was left, corresponding to 29-00 per cent, of V 2 0-, or 
16-29 per cent, of vanadium. 
(2) 1-3155 grm. of crystallized salt from another preparation, precipitated as above 
with sal-ammoniac, gave 0-3680 V 2 O s , corresponding to 27-97 per cent, of V 2 0 5 , or to 
15-71 per cent, of vanadium. 
* This reaction shows that the pyrovanadate of ammonium is not formed by double decomposition, but that 
the meta-salt is precipitated whilst the solution becomes alka line. Thus : — 
Na 4 V 2 0 7 + 4NH 4 Cl= 4Na Cl + 2(NH 4 V0 3 ) + 2NH, + H 2 0. 
2 x 
MD CCCLXX. 
