PotLtoK— On the Extraction of Glucinum from Beryl. 149 
weight, nor was the lustre of the faces of the crystals at all dimmed after 
a fortnight, showing that the crystals were quite permanent in dry air. The 
sulphate is not soluble in anhydrous alcohol, but dissolves freely in 50 per cent. 
alcohol. It decomposes to oxide when ignited for some time over the blowpipe ; 
but a good heat must be maintained for about twenty minutes to ensure complete 
decomposition. ‘The heat of an ordinary Bunsen is not nearly sufficient. 
Many other samples of crystallized sulphate were examined and gave on 
ignition from 14:00 per cent. to 14:20 per cent. of oxide, which would agree exactly 
with the accepted atomic weight of glucinum: but one or two samples gave an 
even lower figure, which is somewhat remarkable, as they were well crystallized ; 
it was noticed that the lowest figures were always obtained from the glucina pre- 
pared by the alcoholic method, and from that portion most soluble in moderately 
strong alcohol. Again, though the oxide as estimated by ignition seemed as a 
rule to bear out the accepted value of the atomic weight when complete analyses 
were made, the ratio of the base to the acid seemed to be subject to considerable 
variation. In the last analysis the acid is too high, but in another sample of very 
beautiful crystals, prepared from the least soluble portion of the glucina obtained 
by the alcoholic method, the following results were obtained. 
Gilucmiasneey ¥ a ve 14:20 
Sulphur trioxide, ... a stat 43°39 
Water, ae me ea ake 42°41 
100-00 
IX. Anuyprous CHLORIDE OF GLUCINUM. 
A number of preliminary experiments showed that the best method of making 
anhydrous chloride of glucinum was to mix the oxide with four times its own weight 
of sugar, and char ina Hessian crucible at a dull red heat; then to place the intimate 
mixture of carbon and glucinum thus formed in a porcelain tube, maintained at a 
bright red heat in a furnace with a Fletcher burner, and pass a steady stream of 
perfectly dry chlorine gas over the heated mass. The chloride formed freely, and 
condensed in beautiful silky needles of a pure white colour, at a point just beyond 
the red-hot zone of the tube; the part next the tube fused to a white mass. In 
the earlier experiments the chloride was yellow, due to the presence of iron; but 
in all the experiments some patches of a light blue colour were noticed, though the 
reason of this was not ascertained. Ground lump-sugar is by far the best reducing 
agent to use, as on charring it makes a very intimate mixture with the oxide, is 
practically free from ash of any kind, and absolutely free from iron, a very minute 
quantity of which utterly ruins the appearance of the chloride, giving it a strong 
yellow colour. It is important that a steady stream of chlorine be maintained, 
