Universidad de Buenos Aires. Facultad de Derecho y.

The American journal of science and arts online

. (page 68 of 100)
Online LibraryUniversidad de Buenos Aires. Facultad de Derecho yThe American journal of science and arts → online text (page 68 of 100)
Font size
QR-code for this ebook


showed 17 p. c. of mother liquor and 36 p. c. of ^^Ta^ SngSig.

8, c. 480 g. of a 1*75 p. c. solution of SaSn were mixed with 400 g.
of a 2 p. c J^aSig solution. There was no change at first, but in the
course of two days it formed a thin jelly. After standing six days it was
draihed, squeezed, and hard pressed, and the cake weighed 26*7 g. This
product contained 21 p. c of mother liquor and 30* p. c. of ^a^Sn^Siit.

These three trials show something of the influence of dilution,
though in the case of c longer standing had also modified the
composition of the coagulum by increasing the amount of stan-
nic oxyd rendered insoluble.

9, a. A 120 g. mixture was compounded so as to contain 18 g. of ^Ta,
Sd2 &k There was no change at first, but in the course of two days it
became a thin opaline jelly. After seven days being hard pressed it
yielded a translucent cake weighing 8*6 g. and containing 37 p. c. net of
fTaodot di,3.

9, 6. A similar mixture was made by adding a saturated solution of
stannate to fused crystals of silicate and evaporating till it contained 40
p. c of solid matter. It remained clear, and when exposed to severe
cold showed no sign of crystallization.

10, A mixture was made containing somewhat over 16 p. c of
JTa, Sn Sia. There was no change for six weeks.

11, a. A mixture was made so as to contain about 16 p. c of
tTa, So Si. There was no change for some days, but in the course of a
month it gelatinized.



Digitized by VjOOQ IC



J. M, Ordway on Waterglass.

11, 5. A saturated solution of stannate was put with as n
]?ra Si fig as would make Sa* Sn Si, and the mixture was eva]
gentle heat till it contained 47'5 p. c of solid matter. It rei
and would not crystallize though set out in the open air di
of the coldest days of winter. The strong liquor hore hoi
apparent change, but when much diluted was coagulated by

Stannate of Potash with Silicate of Potash,

12, A mixture made so as to contain 21 p. c of fa^''^
no visible change in a month.

A mixture containing 19 p. c. of l^sSoSi^i also remainei
for a month.

13, a. 72 g. of a 10 p. c. solution of &Sn were mixed \
12 p. c ^Sis, 80 as to bring together about equal weights <
SiOg. There was no perceptible alteration for a day, but i
it became a thin transparent jelly. After standing a week it '
gradually and then hard pressed. The transparent product
g. and contained 39 p. c. net of ^8804 Si, 3.

13. b. 288 g. of a 2*6 p. c. i^Sn solution were mixed with
p. c. t Sis. It remained apparently unchanged for a week. 1
of the second week it became gelatinous, but no liquor coul
out. At the end of three weeks it had fully coagulated, an(
hard pressed it gave a translucent cake weighing 23 g. Tl
17 p. c. of mother liquor and 40 p. c. of tcSoj Si, 4,

14. 77 g. of a solution containing 16 p. c. of ^Sn, were
33 g. of a 20 p. c. solution of & S'% so as to make in all &:
underwent no noticeable modification for some time, but in
formed a firm, slightly opaline jelly. After standing a week
pressed and yielded a mass weighing 28 g. and containing
of ^4 Sna Si,.

Metastannate of Potash with Potash Waterglasi

15. 20 g. of a 10 p. c. tSia liquor were mixed with 20
tion containing 10 p. c. 6804.4. There was no change foi
but in the course of 17 days it thickened. The hard press<
weighed 6*5 g.

16. 10 g. of a 30 p. c. solution of 6813 were stirred inl
24 p. c. solution of KS05.3. It soon got a little thicker bul
atinize at all.

Weaker solutions of the same metastannate and silicate,
with similar results.

17. 10 g. of a 10 p. c. i^Si3 solution were put with 10 g.
;fi[5D,.5 ; no immediate change. In the course of 10 days i
and when hard pressed it gave a cake weighing 4 g.

18. 20 g. of a 16 p. c. solution of ^Sig were mixed with
p. c. &8dio-0. It continued unchanged a month, but at le
ized.

A comparison of all the experiments from which tl
has been made, brings to light nothing definite or



Digitized by VjOOQ IC



180 /. M. Ordway on Water glass.

the products of the reaction of stannates with silicates. The
strange looking quasi formulas here used to show their compo-
sition by equivalents instead of per centages — like the similar
expressions employed for the precipitates described in Parts III
and IV, — would be altogether preposterous, were they intended
for anything more than compact and readily collatable expres-
sions of the results of analyses. In vain do we search for any
principle that will enable us to assign a rational constitution to
substances which derive their unlimited variety of composi-
tion from mere accidents of dilution, purity, temperature, and
of time allowed for segregation. Many picked instances, like 7,
8 a, and 9, might countenance the supposition of a tendency to
the formation of an exact siliciite of tin SnO, 2Si03, united
with difTerent proportions of alkali. But such hasty generaliza-
tion is checkea by an enlarged view and more particularly by
the special test experiments, 8 a, i, c. All that seems to l)e
predicable of the reactions is that mixtures containing as many
eG|uivalents of alkali as the sum of the equivalents of SnO, and
SiO,, are likely to undergo little or no change at the common
temperature of the air; but when the mixture contains less
alkali, gelatinization will occur in a few hours or days ; and the
curd will be greater in amount according as the strength of the
liquors put together is greater, and as the total proportion of
alkali is less.

The segregated matter retains the alkali with no little force,
for when the air-dried precipitate is washed with water a part
indeed of the alkali is removed, but the greater portion remains
obstinately in combination,

19. Some of the cake of Ida, was reduced to powder and kept over
lumps of caustic soda eighteen days. The dry powder was well washed
with cold water. The air-dried residue amounted to 42 p. c of the
quantity of fresh cake taken, and contained 78*6 p. c of 4E0 .4SnO'.
ISSiO*. Only one- third of the potash had been washed mit.

It is difficult to ascertain whether the fresh undried -cake may
undergo dissociation in any greater degree; for if we attempt
to wash it, though a part settles, the supernatant liquor remains
milky a very long time, and the suspended matter cannot be
separated by filtration, as it readily goes through the pores of
the paper.

After seeing how a deficiency of alkali facilitates the coagula-
tion of a mixture of stannale and silicate, we should hardly ex-
pect to find metastannates so slow in producing any effect on
waterglass. But metastannates are evidently not mere polyacid
stannates, and a higher degree of compatibility is the less sur-
prising when we consider the many points of resemblance be-
tween metastannate of potash and waterglass itseH Both are
ucory^tallizable 4bBd dry to transparent gum-lik:<e masses, indefi-



Digitized by VjOOQ IC



/. M. Ordway on Waierglass.

nitely soluble in water. Both are precipitated by a
partial decomposition, and indeed in both the acid an<
to be in a state of association rather than of strict cb
bination. The following experiments, illustrating
may be compared with the similar trials of silicates
in Part III :

20. 350 g. of a Bolation containing 6 p. c. of t So*.*, we?
860 c c. of alcoliol of sp. gr. 0-840. After two waeliings
the flocculent precipitate was collected in a cloth and haid \
thin translucent, brittle cake, after twelve hours exposun
weighed 24 g. and contained 82 p. c. of ]S:Sn,.4. It was d
parts of water and treated with alcohol as before. The sc
tate when pressed and aired contained 81 p. c. of i^Sn,.,
piecipitate contained 80 p. c. of tijnj.^. The product of
cipitalion showed 81 p. c. of iSnQ.j. The fifth precipitate
p. c. of tsnio.g. And the sixth product was readily soh
and nhowed 72 p. c. of tsn, ,.4.

21. 60 g. of pressed crystals of stannate of potash were
water, so as to make 600 c. c. 200 c. c. of 6*4 p. c. niti
stirred in, and, when the liquor had recovered its transparem
of alcohol were added. After one washing with alcohol ti
precipitate was very gradually squeezed and then hard pr
some hours exposure to the air, tlie cake weighed 27 g. and
p. c.of iSfij.a. It was dissolved in 9 times its weight <
treated with alcohol. The second fiocculent deposit hard
dried in the air, contained 80 p. c of KSoe-a* Theprecipit
peated many times, and finally the tenth product, weighinj
not entirely soluble in water. The soluble part contained
the small insoluble residue had nearly the same composition

Metastannate of potash is also thrown down as 8U<
potash salts ; and here too, as with waterglass, the
the chlorid prove most efficient

22. 26 g. of A solution containing 21 p. c. of £[§04.4 wei
26 g. of a 20 p. c KCI solution. The very bulky deposit
squeezed and subjected to hard pressure, weighed 6*4 g. It
p. c. of mother liquor and 64 p. c. of ftSo,.,, and was read
water.

23. 40 g. of a 9 p. c solution of fe 8n,., mixed with 4
p. c. solution of acetate of potash gave a precipitate that c<
4 g. and contained 68 p. c of ^Soq-i, This was wholl
water.

Carbonate, sulphate, chromate, or nitrate of pot
precipitate after a time when added in inconsiderab
to the metastannate ; but neutral stannate, of whate\
and in whatever quantity, has no effect upon a roetai
lution.

On account of the insolubility of metastannate oi
salts soon disturb a solution of metastannate of potaji



Digitized by VjOOQ IC



182 J. M. Ordway on Water glass.

liquid containing 10 p. c. of 4 805.,, became opaque with its own
weight of a 2 p. c. solution of S^aO. A much less quantity of
sulphate of soaa sufficed to produce the same effect. A soda
salt therefore affords a test of the presence of metastannate in a
stannate of potash solution, but the liquors must not be too con-
centra ted.

Fremy says that stannate of potash is precipitated from its
solution by almost all soluble salts, and even by salts of potash,
soda, and ammonia. He must have operated with liquors that
had been kept too long and had thereoy become contaminated
with metastannate ; for my own experiments afford no confirm-
ation of his statement as far as potash and soda salts are con-
cerned. On the contrary I have not found these salts to have
any effect at all on solutions of fresh and really normal stan-
nates of potash and soda. Stannate of soda when mixed with
a neutral salt, does indeed become turbid by standing some time,
but so does the diluted stannate without any addition.

Indefiniteness of composition pertains to the metastannates in
no less degree than to waterglass. Fremy once assigned the
formula* K0.2Sn30g+6HO, but afterward settled down on
KO . Sn^O, 0+4HO, and thus gives his reasons for the change:*

"Jeprdparais autrefois les metastannates en faisant bouuler
de Tacide metastannique avec des alcalis, et en pr^ipitant ces
dissolutions par I'alcool; j'ai reconnu recemment que pendant
Tebullition, une partie du metastannate se transforme en stan-
nate ; en traitant ensuite la liqueur par I'alcool, je pr&jipitais un
melange on peut-6trc une combinaison de stannate, et de meta-
stannate, etc."

" C'est ainsi que j'avais 6\A conduit a repr^nter.racide meta-
stannique par la formale Suj 0^. Mais je prepare actuellement
des metastannates qui ne pen vent contenir de traces de stan-
nates; aussi leur analyse donne-t-elle toujours un equivalent
d'acide metastannique repr&ent^ par la formule Sn, 0, ,,."

Fremy's later methoa of preparing a metastannate which
should contain no stannate, was to dissolve metastannic acid in
caustic potash and precipitate the combination by adding to the
solution bits of solid caustic potash. The precipitate was dried
by absorption on unglazed porcelain. But it can hardly be pos-
sible by mere absorption to get a soft, pulpy substance, entirely
free from so thick and dense a mother-liquor as the caustic pot-
ash would make. Fremy made five analyses which came out
very nearly alike, as might indeed well happen so long as the
products analyzed were made in just one way and by one indi-
vidual. But the tacit assumption that none of the 10*5 to 10'9
p. c. of potash found, was due to retained mother liquor, is man-
ifestly gratuitous. Nor was there any good reason for suppos-

' Annales de Ch. et de Phys. [8], xii, 476.
*. Id., xxiii, 895.



Digitized by VjOOQ IC



J. M, Ordway on Water glass. 183

ing that cold caustic potash exerts no modifying action on meta-
stannate; for time often brings about in part what heat effects
rapidly and fully, and in fact I have seen an abundant precipi-
tate, produced by moderately strong potash liquor, disappear in
an hour or two with but a very slight elevation of temperature.
Some idea of the diflSculty of separating a mother liquor from a
gum-like precipitate, may be gathered from the experiments on
precipitated waterglass given in part IV. There we had a sub-
stance as capable as metastannate of potash of being reduced to
a very compact form. There we had the advantage of a mother
liquor much thinner and much more easily separable from the
curd. There was also a powerful pressure eflfectually accom-
plishing in a few moments, and consequently without chance of
modification by time, what the contractility of metastannate and
the imbibing power of porcelain could do but slowly. Yet the
precipitated silicates retained, on an average, about thirty per
cent of mother liquor. Due allowance being made for the
greater density of a stanniferous compound, it certainly would
not be making a very high estimate to say that metastannate of
potash thrown down by potash, after as complete a drainage as
possible, must retain not less than ten per cent of strong caustic

r)tash liquor or some three per cent of dry potash in excess,
see no cnance of actually finding out how much of the potash
is combined and how much is free in such a case ; for if we add
any indicator^ it would alter the conditions of the experiment.

Hence, Fremy's formula having no fixed basis must be looked
upon as a rather uncertain approximation to the true composi-
tion of the metastannate made by his method.

According to Rose,* Weber found solid metastannate of pot-
ash to consist of KO. TSnO^ +3H0. But we are not told by
vrhat process it was mkde.

Berzelius says that one part of potash will dissolve sixteen
parts of stannic acid, and these proportions would make about
KO. lOSnO,.

As to matastannates obtained by precipitation with alcohol or
with neutral potash salts, we have found that by varying the
conditions the composition of the products may be made to
range from less than five up to more than seventeen equivalents
of binoxyd of tin to one equivalent of potash. Ana Graham*
has shown that when the alkali is eliminated by neutralization
and dialysis, metastannic acid is itself soluble in pure water.
Hence there is probably no limit to the possible diminution of
potash in the still soluble metastannate.

I have heretofore enunciated the general rule that soluble salts
having as bases, ferric, chromic, stannic, or other oxyds contain-

• Poggendorflf*8 Ann. Ixxy, 16.

* Repertoire de Chimie, Sept^ 1864, p. 184.



Digitized by VjOOQ IC



184 , •/. M. Ordway on Water glass.

ing more than one equivalent of oxygen to one of metal, may
have a large part of the acid withdfrawn and still remain dis-
solved. And now it is interesting to find that when binoxyd
of tin is brought into solution by an alkali, a similar principle
holds good, and the greater portion of the base may be removed
without any immediate permanent precipitation of the acid.
Metastannate of soda, however, is but temporarily soluble, and
is thrown down from solutions of the neutral stannate by long
standing or by boiling. It is also soon deposited firom a solu-
tion in which the alkali of normal stannate has been partly
taken up by a strooger acid. These precipitates are, by no
means, of the same composition, though they often approach
very closely to the proportions required for Na0.5SnO,.

24. Some purified stauDate of Boda that had been dried in the air and
then kept in a well stopped bottle for two years, on being treated with
ten times its weight of cold water, left undissolved one twenty-third of
the tin, comhined with soda enough to make NaO. 3 YSnO^.

The clear liquor, hy standing several weeks, let fall one-seventeenth of
its tin as Na0.4-7Sn02.

25. A ten per cent solution of normal stannate being kept 34 days,
deposited one-twelfth of the oxyd of tin in combination with enough
alkali to make NaO . dSnO,.

26. A ten per cent solution of NaO . SnO^ was boiled a few momenta
and let fall one-sixteenth of its tin with some soda forming NaO . SSnOj*

27. A solution containing five per cent of pure stannate, by boiling
deposited one-seventh of the tin as NaO . S-YSnO,.

28. Boiling a two per cent solution of NaO.SnOj caused the precipi-
tation of over one-third of the tin and enough soda to make NaO . 7Sn02.

29. A one per cent solution of pure stannate of soda required long boil-
ing to make a decided precipitate and the clear liquor filtered out of the
bulky product very slowly. The well drained gelatinous residue was
soluble in water and consisted of NaO.TsSnO^. It contained over one-
half of the tin.

30. A solution containing I'l p. c. of NaO.SnO^ and 3 p. c. of NaCl,
being boiled five minutes, gave a dense, opaque precipitate very easy to
collect, drain and press. It contained one-fourth of the tin and consisted
of NaO.a-TSnOa.

The addition of chlorid of sodium and subsequent boiling causes a
precipitation in a solution containing no more than 0*1 p. c. of normal
stannate of soda.

So far as is known at present, there is nothing to show that
any one of the metastannates more than another is entitled to
rank as a definite chemical compound. None crystallizes ; none
is a product invariably coming the same by several different
ways of formation; none exhibits a plain analogy to any un-
doubted exact chemical combination ; nor is there any point up
to which the variation gradually decreases, and beyond which
it gradually increases. There being then nothing certain on



Digitized by VjOOQ IC



/. M. Oidway on Water glass. 186

which a formula may be based, it seems hardly right to take
any one chance product, and, rejecting the odd parts of equiva-
lents, to set down that product with its analysis so amended, as
the true metastannate of potash or of soda. The composition
of substances can not always be squared with exact atomic pro-
portions, for there is a class of indefinites in which one of the
constituents may admit of a large increase or diminution with^
out apparent alteration in the character of the compound. As
the world was very slow to comprehend Wenzel's proposition
that bodies must unite in fixed and invariable proportions, so
has it since been quite as slow to learn that there are cases in
which this great principle does not hold good. Wenzel says:^
"Dass eitie jede Verbindung der Kdrper, eine bestimmte und
unveranderlich bleibende Abmessung haben muss, die ohne
aiisserlich mitwiirkende Ursachen weder grosser noch geringer
werden kann, weil sonst auch nichts gewisses aus ihrer Ver-
gleichung bestimrat werden konnto, ist schon an sich klar. Es
folgt daher nothwendig, dass eine jede mcJgliche Verbindung
zweyer Korper, mit jeder andern bestandig in dem genauesten
Vernaltnisse stehet, und dieses Verhaltniss driicket den Grad
der Verbindung aus." That the constitution of bodies is deter-
minate and unalterable, is hardly self evident, and indeed it is
a matter to be proved by the balance rather than by mere rea-
soning. It is true that the confirmatory instances have been
found to be numberless. Still a great many do not necessarily
make all, and a single negative example is sufficient to disprove
absolute universality. Wenzel himself, while laying the sure
foundations of chemical science, fell in with some of the indefi-
nites and thus unwittingly furnished some negative instances.
He proceeded to determine the equivalents by neutralizing the
acids with different bases, and worked correctly so long as those
bases were protoxyds. But such a method is as little applicable
to alumina and similar peroxyds as it is to bone earth, — **Elfen
beinerde," — which he reckoned among the simple bases. So
after attempting to saturate nitric acid with alumina, he con-
cluded that* "Das Verhaltniss der Alaunerde zum starksten
Salpetersauern, ist also beynahe .... wie 349 : 240." Instead
of forming the normal nitrate Wenzel in this case must have
got the soluble basic 3AI3O3 . 2X0^, and stopped at a supposed
but not real limit of solubility.

Cases of indefinite combination are not unlike those of solu-
tion in which such things as gum or albumen are concerned.
These may unite with water in all proportions, while most crys-
tallizable substances dissolve, at any given temperature, until
the liquid contains a certain percentage of solid matter, — a per-
centage which is exactly definable for each particular salt. In

^ Lehro vod der Verwandschaft der Eorper. Dresden, 1782. p. 4.
• Idem. p. 118.
Am. Joub. Soi.— Sboond Sbries, Vol. XL, No. 119.— Sept., 1866.
24

Digitized by VjOOQ IC



IM . J. M. Ordway on Water glass,

the pbenomenon of fusion too, though almost all bodies liquefy
at a temperature which is invariable and exact for each, yet
there are a few substances that pass through an intermediate
state of softening before fully meltinK* In fusion as well as in
solution, abrupt transition from the solid to the fluid state is the
general rule ; and yet, as everybody allows, it is not the univer-
sal law. And why should we be over-reluctant to admit the
possibility of some instances of chemical union in indefinite pro-
portions r

Normal Silicate of Soda. — The only alkaline silicate certainly
known to be crystallizable, contains two Berzelian equivalents
of silicic acid, SiO,, to three equivalents of soda. Bespecting
the amount of water in this salt, analysts do not agree; and as
the several observers have examined crystals obtained by unlike
methods, there may possibly be three or more hydrates produci-
ble under different conditions of temperature, concentration, and
purity.

Fritzsche* produced a salt in the form of square prisma, by
dissolving in caustic soda lye a (quantity of silica equal to that
of the drv soda present, and allowing the solution to rest several
days. The bruised crystals freed from mother liquor by pres-
sure between folds of absorbent paper, gave him amounts of
silica and chlorid of sodium corresponding to NaO.SiOj .9H0.
These crystals melted at about 40° 0. Kept over sulphuric
acid they in time effloresced (verwittert) to the center.

Fritzsche says that he once chanced to obtain globular masses
studded with minute crvstals and having the com{)Osition NaO.
SiOj . 6H0 ; but he had. never succeeded in determining the con-
ditions of their formation.

Yorke" fused 23 parts of sand with 64 parts of dry carbon-
ate of soda, dissolved the resulting mass in water, ana exposed
the solution to slow evaporation in vacuo, over sulphuric acid.
The rough salt was recrystallized in vacuo, to get ria of the ex-
cess of carbonate, and the pure product was found to have the
composition NaO.SiO, .7H0. He also obtained crystals of the
same composition from a solution of silica in caustic soda.

Yorke round these crystals to part with all their water, except
a fraction of one per cent, by exposure to a heat of 149° C.

Hausmann" says that in the purification of rough soda, the
mother liquor has often yielded him, in large quantity (manch-
mal centnerweise), rhombohedral crystals of silicate, permanent
in the air, and having the the composition NaO.SiO, .8H0.

Ammon " also obtained the octonydrate in the form of mono-
clinic crystals, by dissolving silica in soda lye. They fused at
45^0.

• Popg. Ann^ xliu, 185. »• PhiL Tran»^ 1857, p. 588.

" Eidmaun's Journal, xii, 294.

" Will and Kopp^s Juhresbericbt, 1862, p. 188.



Digitized by VjOOQ IC



/. M. Ordway on Water glass. 187

I have found the same melting point, and have made several
analyses which invariably go to confirm the formula given by
Haasmann and by Ammon. Thus in one case the pure, care-
fully dried crystals, by suitable treatment with nitric acid, gave
22-74 p. c. of SiO,, and nitrate corresponding to 28-66 p. c. of
^aO. Another portion of the same sample being heated with
fused bichromate of potash, lost 54*18 p. a

The percentages answering to NaO.SiO, .8H0, would be
22 55 of SiO„ 23-81 of NaO, and 54-14 of BO.



Online LibraryUniversidad de Buenos Aires. Facultad de Derecho yThe American journal of science and arts → online text (page 68 of 100)