arch 35,155} Defence of Turner's Method of Detecting Boracic Acid. 12
Coos dn OLE 0% 2 sw 1 led to a suspicion of the presence of sulphur compounds
THE CHE MI CAL NEW S. of lower oxidation in the acid sulphite, and the salt was
" examined and found to contain hyposulphite.
In order to ascertain whether the occurrence of this
impurity was the result of accident, I lately procured from
six of the best known laboratory furnishers a small quantity
 of their pure acid sodium sulphite, as sold for the
reduction of ferric salts in the “ volumetric estimation of
iron.” Upon the arrival of the samples, they differed considerably
 in appearance, and ranged in colour from pale
straw to white. The percentage quantity of matter insoluble
 in water was very small in each case, ranging from
0°14 to 0°06. When tested for hyposulphite only one
sample was found to be free; it was from Messrs, Townson
 and Mercer. For the others, no two specimens were
alike in the quantity contained ; there was, moreover, a
very wide difference between the potential of reduction of
the best sample and the worst. When tried upon solutions
 of ferric chloride of known strength under exadly
similar conditions, 40 parts of the best sample were equal
in power to 100 of the worst. Now it seems to me that
when a certain salt is sold as pure and fit for analytical
work, some effort ought to be made by the firm selling the
substance to ensure its being as near as possible the
article it professes to be; because often in an aQive
laboratory there is little time to spare for more than the
application of a few simple tests to ascertain the comparative
 purity of reagents, and very considerable reliance
is often—too often in fact—placed upon the name and
fame of the firm supplying the material, and again it is
especially necessary that the chemical reagents supplied
to those who are struggling with the difficulties of a
chemical education should really be pure in nature as well
as in name. Imagine the embarrassment of a student on
finding a coloured precipitate formed in a solution of say
copper, arsenic, and ferric iron by the use of * pure”
sodium acid sulphite, which he had added for the sole
purpose, and with the one expected result of simply
lowering the condition of the arsenic and iron previous to
passing H,S. It is difficult to conceive how the variation
here dealt with occurs, for the methods of preparation of
the ‘“bisulphite’ seem to be held within very narrow
limits; it is a question, however, which deserves an
answer, and I hope shortly to be able to communicate
further with you upon this subject. Meanwhile I must
apologise for the meagre character of this note, and claim
indulgence for it solely upon the ground that there is
3ood in everything which tends to equalise the conditions
under which similar work is performed by different
analysts.

IN the July number of the Yournal of the Chemical Society
ii., p. 115, 1877, there appeared a short abstract of a paper
by E. H. Jenkins (Fourn. Prak. Chem. [2], xiii., 237—230)
which stated that the presence of silicic acid does not
interfere with the determination of phosphoric acid by
means of ammonic nitro-molybdate, and that it is there.
fore unnecessary to remove it. As this seemed contrary
to all experience, and thinking that it ought to be confirmed
 or corrected, I requested two of my students to
make some estimations of phosphoric acid by this process
in presence of different weights of silicic acid, and also in
its absence: The following are the results :—
a5 c.c. of a solution of sodic phosphate containing
0025 grm. P,05 were employed in each experiment.
Ammonic nitro-molybdate was added, the liquid allowed
to stand for eighteen hours, the precipitate washed, dissolved
 in ammonia, filtered, magnesia mixture added to
the filtrate, and the precipitate weighed as magnesic
pyro-phosphate. The silicic acid was in the form of sodic
silicate.
First Series, by Mr. Watanabe.
Co Mg,P,0,. P,0,.
1. No silica 0°0390 grm. = 0'0249 grm.
2. No silica . 00395 , = 0°0253 ,,
3. 03935 grm. SiO, 0'0455 4 = 00201 ,,
4. 07870 » 0'0505 ,, = 00323 ,,
5. 1°1805 ” 00555 4 = 0'0355
Second Series, by Mr. Takamats.
1. No silica 0'0392 grm. = 0'0251 ,,
2. 0223 grm. SiO, 0'040g , = 0'0262 ,,
3.0553 00429 ,, = 00274 ,,
4. I'106 » 0°045X , = 0°0288 ,,
In both series it will be observed that the weight of the
magnesic pyro-phosphate increases with the amount of
silicic acid present, and therefore the apparent amount of
phosphoric acid increases likewise. Although there is
not a perfect agreement between the two sets of experiments
 as to the increase in weight of the magnesian pre.
cipitate for a given weight of silicic acid present in the
solution, still, I think, it is sufficient to show that the
silica ought to be completely removed before estimating
the phosphoric acid. I trust you will oblige me by giving
publicity to this correction.
University of Tokio, Japan,
February 7, 1877.

g——

A DEFENCE OF TURNER'S METHOD OF
DETECTING BORACIC ACID.*
By C. LE NEVE FOSTER, B.A, D.SC., F.G.S.

A DEFENCE of Turner's method+ seems at first sight to be
Juite unnecessary, for his test has been used over and
over again, for upwards of fifty years, and, as far as I am
aware, its usefulness has never been questioned till quite
lately. Great was my surprise a few days ago to read a
a paragraph “On the uselessness of Turner's flux as
applied to the Detection of Boracic Acid.” Professor
Chapman speaking of Turner's method, says :-—* This
test is much quoted in blowpipe books, and works on
chemical analysis generally; but it is altogether superfluous.
 With borate of soda it fails entirely, or yields
very unsatisfadtory results; and although it answers for
* A paper read before the Mineralogical Society, December 18th,
oe consistsin heating on Pt wire before the blowpipe the powdered
mineral mixed with a lux made of 44 parts of bisulphate of potash,
and cne part of finely powdered fluor spar. If boric acid is present
i ‘el .
the or some. re fonctions, by Prof. E. J. Chapman, Canadian
Yournal, October, 1876.

ON THE DIFFERENCE IN QUALITY OF SOME
SAMPLES OF SO-CALLED PURE SODIUM
ACID SULPHITE.
By W. F. K. STOCK, F.C.S.

SoME time ago I received a supply of sodium acid sul.
phite, which was labelled ** Pure for Analysis,” and ip
using it for the reduction of acid solutions of ferric salts in
the separation of phosphoric acid, I found that after neu.
tralising with sodium carbonate the reduced solution, the
addition of sodium acetate produced a precipitate which,
though light red at first, became deeper and deeper upor,
boiling, until it had acquired a dirty olive-grey tint. This