Rodolfo Amedeo Lanciani.

The American journal of science and arts online

. (page 24 of 109)
Online LibraryRodolfo Amedeo LancianiThe American journal of science and arts → online text (page 24 of 109)
Font size
QR-code for this ebook

OojOj which dissolves in €5aH4 0, with a brown, and in
-G3H3O4 with a green color. This observation is of undoubted
value. He says also that on mixing the neutral solutions, the
precipitate is slow in forming, and forms by an absorption of
O, as is shown by dipping the connecting tube under a cylinder
of air or oxygen. Erdmann confines himself to an examination
of this last observation of Stromeyer's and finds that the
" neutral salt " can be formed in an atmosphere of pure -GO,,
and that therefore there can be no absorption of O in the case.
Braun finds that KHO acts very little upon the salt, merely
changing' its color from yellow to a dirty yellowish-green.
NaHOjBaHtOs, -CaH|€t3 separate at a boiling neat the brownish-
black hydrated GojOs- This is confirmatory of Stromeyer's
observation. Carbonate of silver heated with water and
Fischer's salt gave with separation of the hydrated CJojO,,
crystals of what from Braun's description were probably
AgNOj. KCy did not decompose it

* Zeitsch. AnaL Oh., Tii, 313. f Ghemie der Jetsteeit, p. 414.

% Should be 608.

Digitized by VjOOQIC

& R SadOer m Fischer's Salt 19J

My own experiments directed toward the solution of the
question of OoO or €k)|Os weieas follows : I found Erdmann*s
observations as to the formation of the " neutral salt" in an
atmosphere of €JOa correct I added by means of a fiinnel
tube a solution of KNOj boiled to expel the air, to a boiled
solution of Co Clj, placed at the bottom of a partially closed
vessel, into which a strong stream of €0, had been passing for
some time. The salt began nevertheless to form slowly.
Wishing to sift the matter more thoroughly, I made a series of
experiments in eudiometer tubes oVer mercury. Into a eudi-
ometer tube filled with mercury and perfectly free from air, I
threw up a boiled solution of CoClg and tnen one of KNO,.
The salt began to form soon at the juncture of the two liquids
and then deposited steadily though slowly. After a number
of hours not the slightest bubble of air was to be seen. The
absence of all chance of oxydation here, either from air or
excess of nitrous acid, shows conclusively the protoxyd nature
of this "neutral salt" The "acid salt" was formed in the
same manner — ^first a boiled solution of CJoClj, then acetic
acid, and then a boiled KNO3 solution. A rapid evolution of
gas ensues and the salt forms and falls in a layer on the top of
the mercurial column. In another instance a small crystal of
•CoClj was sent up and then acetic acid and then KNOj. The
formation of the salt was almost instantaneous and the evolu-
tion of ^s exceedingly rapid. Before drawing any conclusions
fix)m this, however, tne action of acetic acid on KNO,, out of
access of air, was to be studied. Gmelin* states that, when
treated with acid out of access of air, nitric oxyd is evolved
while the liquid takes up NOj and NjOj. This evolution was
found to taKe place readily over the mercuiy and the gas
answered to the test of ferrous sulphate and was colored red
on admission of air, like the gas evolved in the formation of
the salt The question now is whether the liberated O is taken
up in the formation of the salt to oxydize the €oO, to convert
the excess of nitrite into nitrate. The question cannot be
definitely settled, until we have some delicate test for a small
quantity of nitrate in the presence of a large quantity of
nitrite. But there is one thing which I regard as significant
Erdmann made one of his preparations of wie " acid salt," by-
filtering the mixture of neutral solutions into pure acetic acii
Now the greater the excess of acid, the quicker would be the
conversion of nitrite into nitrate, and we should look for a very
small amount of the double nitrite salt Yet it formed as
readily, and analysis proved it to possess the same constitution
as the other preparations. The inference from this is, that the

♦ Cav. Soc. ed. vol ii, p. 382.

Digitized by VjOOQIC

192 S. P. Sadder on Fischer's Salt

sesquioxyd-fonning power of the €o is sufficient to enable it
to take liberated oxygen.

Recourse was now nad to analysis as a means of determin-
ing the question. Six distinct preparations of the salt in aU
were made and analyzed. In preparing the salt, the distinction
made by Erdmann m r^ard to the formation of different com-
pounds in neutral and acid solutions, was recognized and the
salt was invariably formed in a solution, stronmy acidified by
acetic acid in the beginning, and kept so untu the end by the
addition of acid, if necessary. The salt of cobalt used was
the chlorid, and, in all but the first two preparations, the solu-
tion was boiled to expel air and made acid. A strong solution of
KNO was then adaed and the mixture was left to stand : 6
to 12 hours generally proved sufficient, and it was then filtered
and washed, accordiSag to Erdmann's direction, with a solution
of potassic acetate of about 10 per cent which was displaced
by 80 per cent alcohol The salt was then dried, first over the
water-bath and then in a water-oven exactly at 100°. I saw no
evidence of a decomposition at this temperature alluded to by

The results of full analyses of these six preparations warrant
me I think in presenting the following conclusions : —

1st That Fischer's salt is a Tri'potassic-OobaUiC'Nitritej its es-
sential formula being €o,0„8N,e3-|-3(K,0, N,0,)+Aq, or

2nd. That it can be formed with 4H,e, 3H,0, 2H,e, H,e,
or anhydrous, according to the d^ree of concentration of the
solutions used, passing in color from a light yellow to a dark

8rd. Tnat in consequence of such dependence, we can, in
in most preparations, fix no absolute point, but are liable to
have a mixture of salts of different degrees of hydration.

The analytical methods used in the analyses of the salt were
the following : —

The €5o was determined first together with the K, as a double
sulphate having the invariable ' constitution €o, K,5S0^.
This sulphate has a distinct fdsing-point, and affords an accu-
rate means of determination. It will be discussed fiBJi;her on.
This is then dissolved, the solution made ammoniacal and
brought to boiling, when the €o is thrown down as "GoS. This
is roasted and, after treatment with aqua regia and sulphuric
acid, is weighed as neutral OoSO-. Or the €Jo was tnrown
down fix>m the solution of the double sulphates as peroxyd by
acetate of soda and chlorine, and then reduced by hydrogen to
the metallic stata

The potash is then determined by difference.

Digitized by VjOOQIC

& p. SadOer on Fisdier's Salt 193

The nitrogen was detennined, either together with the H,0
by Gibbs's modification of Bunsen's method, which will be again
alluded to, or directly by volume with the Sprengel-pump.

The H,0 either as above, or by combustion with -Gu in a
stream of €0,.

The analytical data and results are appended.

Preparation No. 1.

•2011 gr. salt gave -1732 gr. €o,Ke6^04=44-75 p. c. €oe4.K20
gave also -0656 gr. £o8e^==U'18 p.c, €oe=17-47 p.c.

•1626 gr. salt gave -1403 gr. ■eo2Ke5Se^z=44-83 p. c. ^oO-fKaO
gave also ^0529 gr. €o&e4=15-74 p.c. €oe=17-42 p.c.

•5147 gr. salt gave 112*7 cc. nitrogen at 15*25^ and 483*2 mm. pres.

= 16-56 p. c. N=44-95 p. c. N^Og.
•9410 gr. salt gave 167*6 cc. nitrogen at 15*25 and 601*6 mm. pree.

z=16^76 p.c. N=45-48 p.c. N^O,.
11131 gr. salt gave -0793 gr. Hje=7^pO p. c. H^O.
1-3418 " " -0937 " " 6^98 " "

Preparation No. 2.

•1082 gr. salt gave -0955 gr, €o2Ke5Se4=46-86 jp. c. €©0+ K,0
gave also -0408 gr. €o&04=16^16 p. c. •GoO=l7*88 p. c.

•1222 gr. salt gave -1074 gr. €02X^5804=45-66 p.c. GoO+KaO
gave 5so -0455 gr. €o804=15-96 p.c. €oO=17*66 p.c.

•4192 gr. salt gave 102*83 cc nitrogen at 18*25° and 462*6 mm.
pre8.=17'57 p.c. N.=47^70 p. c. NjOj.

•4376 gr. salt gave

•6328 " " -0346 gr. H20=5*46 p. c. H,e.

•5494 « « -0296 gr. H, 0=5*39 p. c. H^O.

Preparation No. 8.

•e215 gr. salt gave -5682 gr. €02K«5SO4 =47*50 p. c. OoO+K-O
gave also '0797 gr. €o=12-82 p. c. €o=18-04 p. c. Oo«0-.
•58^8 gr. salt gave *5349 gr. €o2Ke5SO4=47*60 p. c. -GoO+K^O
gave ^o ^0740 gr. €o=12-68 p. c. €o=17'83 p. c. Go^Oo.
•4451 gr. salt gave -4088 gr. €o,Ke5S04=47*71 p.c. €oO+K«0
gave also '0574 gr. ■eo=12*89 p. c. €0=18*14 p. c. Oo^O-.
•5796 gr. salt Mve -5329 gr. €o2Ke5&04=47*77 p. c. OoO+K^O
gave wso *0735 gr. €o=12*68 p. c. €o=17*83 p. a ^OnO..
With 1*6634 gr. salt, eombustion-tabe lost •8616 gr., of which
€a€l2 tube took up '0710 gr.=4*27 p. c H^O leaving -2906
gr.=17*47 p. c N.=47-42 p. c. N jO,,.
With 1*6123 gr. salt, combustion-tube lost •8302 gr., of which
€a€lj tube took up -0610 gr=:4^03 p. c. H^O leaving '2692
gT.=17*80 p. c N.=48-32 p. c. N^Oj.
Aat JouB. Sci.— Second Sbbibs, Vol. XLIX, No. 146.— Mabch, 1870.

Digitized by VjOOQ IC

194 & P. Sadder on Fischer's SaJt

Preparation No. 4.

6677 gr. salt gave -6073 gr, €02Ke6Se4=:47-26 p.a ^oO+KaO

fave also '2258 gr. ■eoS04=16-36 p. c. €00=1811 p. c.
•3590 gr. salt gave -8264 gr. €o2Ka5Se^=47-24 p. c. ^oO+K^O
gave also '1226 gr. €o&04=16-62 p.c. €oO=18'29 p. c.
With 1*1138 gr. salt, combustion-tube lost '2461 gr., of which
€a€Jl2 tube took up •0439=:3-94 p. c. H^O leaving -2022 gr.
z=18-16 p. c. N.=49-27 p. c. NjOj.
With 1*2494 gr. salt, combustion-tube lost '2674 gr., of which
t;a€la tube took up '0407=8*26 p. c. H,0, leaving -2267 gr.
= 18-14 p. c. N.=49*26 p. c. N^Og.
With 1*1587 gr. salt, combustion-tube lost '2533 gr., of which
•Ga-ei^ tube took up •0432=3*73 p. c. HgO leaving '2101 gr.
=±18*13 p.a N.=49*20 p. c. NjOa.

PreparaMon No, 5.

•2896 gr. salt gave •2633 gr. eOaK^SSO. =47*26 p. c. CoO+KaO

gave also ^0382 €o=13*20 p. c. €o=18"56 p. c. -GoaO.*?.
•3998 gr. salt gave *3644 gr. €o2Kg5S04=47*42 p. c CoO-fK^O

gave also -0531 -eo=13*30 p.c. •€o=18'70 p.c. -eojOg.
•6747 gr. salt gave -5243 gr. eo^Ke 5^04=47*40 p. c €oO+K„e
gave also '0769 gr. €o=13'38 p. c. Co=18*82 p. c. Cog^^^.
•5959 gr. salt gave *5443 gr. €o2Ke6S04=47*46 p. a CoO+KjO
gave also *0768 gr. -00=12*89 p. c •00=18*14 p. c. OoaO..
With 1*2557 gr. salt, combustion-tube lost '2758 gr., of which
•ea€lo tube took up •0459=3*66 p. c. H^O leaving -2299 gr.
=18*31 p.c. N.=49*69 p.c N-O3.
With 1^1464 gr. salt, combustion-tube lost ^2556 gr., of which
OaOlj tube took up •0479=4^18 p. c, H^O leaving -2077 gr.
=18^12 p. c. N.=49*18 p. c. NjO,.

Preparation No. 6.

•8154 gr. salt gave ^7644 gr. €o2Ke5&04=48^7l p. c. OoO+KgO

gave also -2823 gr. €oS04=16^75 p.c. €00=18*64 p.c.

•5463 gr. salt gave -5110 gr. Co^KeeSO^ =48^69 p. c. CoO+K^O

gave also '1885 gr. €oS04=16*73 p.c. €oO=18*51 p.c

•6806 gr. salt gave -6373 gr. €0aKg5S04=48-66 p. c. OoO+K^e

gave also •2358 gr. €oS04=16'77 p.c. €oO=18*56 p.c.

€0-0 3.
•3712 gr. salt gave -3480 gr. Co^K^eSO^ =48^71 p.c €oO+K,e

gave also -1297 gr. €oSO4=16-90 p. c €oO=18-71 p. c

With 1^5460 gr. salt, combustion-tube lost '3033 gr., of which
€a€l2 tube took up •0116=0*76 p.c HjO leaving •2917 gr.
=18-87 p. c N=51-21 p.c N^Oj.

Digitized by VjOOQ IC

S. p. Sadtler on Fischer's Salt


Prep. No. 1.

Tbeor. p. oto. Found p. cti.

-eo-Og- 17-00 - 17-46

KjO— 28-94 29-00

N.e«— 46-69 46-21

H,e— 7-37 6-99

Prep. No. 8.

Theor. p. cts. Found p. cts.

CoaO,— 17-65

K,e— 30-04

NaOa— 48-48

HoO— 3-88


Prep. 2^. 2.

Theor. p. cti Found p. cts.

17-32 17-77

29-48 29-70

47-57 47-22

6-63 6-43

Prep. No. 4 and 5.

Theor. p. ctt.





Foond p. cts.

18-20 18-66

30-81 30-61

49-24 49-43

3-64t 3-92t

Prqp. No. 6.

Theor. p. ott. Found p. cti.

Co-Oj— 18-35 18-58

KaO— 31-24 31-90

N^Oa— 50-41 51-21

HjO- 0-— 0-75

It will be seen that, while the general coincidence is well sus-
tained, there are several decided deviations. We can only call
attention to certain facts in explanation 1st We probably have
in some of these preparations, mixtures of the salts of dmerent
degrees of hydration. 2nd. The salt cannot be purified by
re-crystallization. 3r no crys-
talline character. The appearances were indicative of fiie
successive formation of three protoxvd oompounda This how-
ever, I found was not the case. The black precipitate under
the microscope was seen to consist of beautifully defined cubes
of a very dark green color, and the green precipitate was seen
to consist of similar cubes, though much smaller. Analysis
afterward proved them to be the same salt in different states
of aggregation- On adding a warm concentrated solution of
KNO3, however, to a warm concentrated solution of €JoClt
nothing is formed but a flocculent yellow precipitate, which ap-
pears to be the same as that which I formed over mercury in the
eudiometer tube. The formulas of these salts appear to be as
* Rojal Swedish Acad. Trans., 1864.

Digitized by VjOOQ IC

S. p. Sadtler on Fischer's Salt 201

follows. For the black or green salt 2(€o2Ne2) + 2(KNe^) + HjO
and for the yellow salt €o2Nre2+2(KNe2)-fHaO, and the first
may be termed a potassw-dicobcUtous nitrite and the second a
potassio-moruhcobaltous nitrite. Appended are the analytical

Prq>, No. 1. — Black Salt, very crystalline.

•5421 gr. salt gave -5421 gr. eOaK^SSO^ =50-48 p.c. CoO+K^O
also gave 3425 gr. €o&O-=30'67 p. c €oO.

•7341 gr. salt gave -7361 gr. €o2K23^e4=60-57 p.c. eoO+KaO
also gave •4662 gr. €o&O4=30-73 p. c. €o0.

Prep, No. 2. — Oreen Salt (had not been washed for analysis.

•3003 gr. salt gave -3063 gr. €oaK28S04z=51-44 p.c. €oO+K,0
also gave -1846 gr. ■GoS04=29-74 p. c. €oO.

•3221 gr. salt gave -3286 gr. €02Ka3SO4=61-45 p.c. CoO+KaO
also gave -1971 gr. €0^04= J9-61 p. c. CoO.

Prep. No, 3. — Black and Oreen Salts mixed, very crystallina

•6942 gr. salt gave -6978 gr. -GoJK 8Se*=50'70 p. c. €oe+aKe
also gave '4410 gr. €oSO 4=^30-74 p. c €oO.

•4713 gr. salt gave ^4730 gr. ■Go8Kj.3SO4=50-62 p.c. €oe=Kae
also gave -3004 gr. €oS64=30-84 p. c. €o0.

Prep. No. L— Yellow Salt

•2702 gr. salt gave -2631 gr. €oKa2Se4=50-06 p. c. -GoO+Kae
also gave 1235 gr. £Jo&04=22^12 p. 0. €oO.

•4615 gr. salt gave '4497 gr. €oKa2Se4=50-08 p. c. -GoO+KaO
also gave -2103 gr. €oSe*=2206 p. a €oO.





Found p. e. etB,

Tbeor. p. ett.

Found p. cts.



















- 3*67


In looking over the analytical data of these analyses of pro-
toxyd salts, we notice a striking similarity between the weights
of salt taken and weights of double sulphates obtained. The
explanation is very simple, and yet I believe there is involved
in it a feet of importance. The similari^ of weights is due
to the similarity of atomic weights of NaO, and SO, (76 and
80) with which the bases are combined. Now in none of the
very many determinations made in the " acid salts," do we find
such a similarity. On the contrary we find the weight of double
sulphates to be invariably less than the weight of salt taken and
in a fixed ratio. The explanation of this is that ■eoa083N30s by
digestion with H2SO is converted into two atoms of neutral
€o&e* and becomes 2(€oS04). 2€oO,2Na08 simply assumes

Digitized by VjOOQ IC

202 & P. Sadtler on Fischer's Salt

SOg instead of NaOj and becomes likewise 2(-eo&04). The
first loses and atom of NO2 which the second does not We
have here a strong evidence of the totally diflferent habitus of
the €0 atom in the " neutral " and in the " acid " salts.

Having now made a complete statement of all mv analytical
results in the case of Fischer's salt proper, as well as in this
series of analogous and related compounds, which I have
described in this connection, and noticed the conclusions to be
drawn fix)m them, before summing up mv argument, I will
examine how the results of Stromeyer, Erdmann and Braun
compare with my own. In the first case I think the last argu-
ment drawn from the similarity of weights of double nitrite and
double sulphate in the protoxyd salts and the difference of
weights of double nitrite and double sulphate in the sesquioxyd
salts, effectually precludes our assigning a formula like that of
Stromeyer, in which 2(N208) combmes with €o«Oii, to the nor-
mal Fischer s salt The salt Stromeyer analyzed was probably
a mixture of " acid " and " neutral " salts, as he speaks of adding
"a little acetic acid." It is easily seen that a mixture of the
" acid salt " and the yellow " neutral salt " would give figures
approximating to his. Erdmann^s results correspond in the
main to the analysis of my second preparation. The only differ-
ence then is that I have introduced a generalization, which I
think my figures fully justify. With Braun's resiilts, I am
hardly able to make a comparison, as he made his preparation
in a manner totally different from that followed by either Erd-
mann or myself His formulas, as might be inferred fix^m my
statement of them in the beginning of this paper, I am still less
able to accept

To sum up briefljr, the reasons for regarding Fischer's salt as
a potassio-coDaltic nitrite, having the essential formula
€o,6Ne8+6(KNe )-|-aq. are—

1st The liberation of O at the moment of its formation and
the rapidity with which, under these circumstances, the salt
forms, would seem to justify the assumption.

2nd. The analyses of six distinct preparations sustain it

3rd. The existence of corresponding sodic and ammonic salts
is proved by analysis.

4tli. The formation of substitution compounds exactly analo-
gous to well known salts is proved by anaiysia

5th. The analogy of Fischer's salt to double cyanids, chlorids
and nitrites proved to contain a similar hexatomic atom.

6th. The difference of ratio between the weights in the pro-
toxyd and sesquioxyd salts is marked and constant throughout

Accepting these proofs, we will sum up with equal brevity
the formulas of the salts analyzed on this view : —

Digitized by VjOOQ IC

JS. p. SadOer on Fiaclier's Salt 208

Tri-potassio-cobaltic nitrite, €Jot6NOa+6(KN02)+aq.

Di-sodio-cobaltic nitrite, €0a6NO2H-4(NaNO2)+HaO.

Tri-8odio-cobaltic nitrite, €o26Nea4-6(NaN08)4-H20.

Di-ammonio-cobaltic nitrite, €o86NOa-f-4 ( N ll4N02)-|-2H20.

Tri-ammonio-cobaltic nitrite, €oa6NOa+6(NH4N02)-|-2H80.

Lut eocobaltio-cobaltic nitrite, €o26N Oa+CJoa 1 2NH8, 6 N Oa+HaO.
Roseocobaltio-cobaltic nitrite, -eo26NOj+€oalONHa, GNOa+HaO.
Xanthocobaltio-cobaltic nitrite,

Potassio-dicobaltous nitrite, 2(-eo2NOa)+2(KNOa)+H80.

Potassio-monocobaltons nitrite, €oaNOa-|-2(KNOa)4-HaO.

These it will be seen are capable of ready conversion into atomic
formulas, thus: €oa^^[K(NOa)2]6 and £roa"'[(Na(NOa)2)a (NOa)^
and so on-

In the course of my examination of these salts, I had an
opportunity of comparmg the relative merits of the diflferent
analytical methods employed. A brief statement in regard to
them may be of some value.

1st Determination of -Go and K. The double sulphate
method already alluded to, was first proposed by Gibbs and
Genth,* and 1 find it to give extremely close results. Sul-
phuric acid is added to the weighed portion of the salt in a
pjorcelain crucible, which is then heateid carefully, either by a
ring gas burner from above, or by radiation while supported in
a sheet-iron crucibla After the nitrous acid is all driven off
with most of the excess of sulphuric acid, the sulphates begin
to fuse and as the last trace of sulphuric acid goes off, thev
pass into calm fusion. They will then bear a very consid-
erable increase of heat without ferther decomposition. I
found this method capable of extension to both the soda salts,
as well as to the protoxyd salts, the double sulphates formed
having of course a constitution depending upon the proportion
of the bases in the original salta The sSda sulphates aid not
appear, however, to possess as perfect a fusibility as the double
potash sulphate fix)m Fischer's salt, while the double sulphates
from the protoxyd potash salts were also less fusible, in propor-
tion to their greater ratio of -Co, I give an instance in proof
of the accuracy of this method. Four determinations of
€oO+K20 were made inpreparation No. 6 of Fischer's salt in
the manner described. Tne percentages of double oxyds were
48-71, 48-69, 48-66, 48-71. This method was tried by Gauhef
without obtaining good results. Several points are to oe noticed
however. The salt in which Gauhe sought to verify the compo-
sition of the double sulphate was plaimy a mixture of " acid "
and " neutral " salt He says, " after some days small amounts
deposited." This was before Erdmann's paper had appeared.

* Besearches on Ammonia oobalt bases, p. 49.
f Zeitsch. AnaL Ch., iv, p. 56.

Digitized by VjOOQIC

204 S. P. Sadtler on Fischer's Salt

Again, Qauhe will find tliat Gibbs and Genth never even men-
tioned the use of (NH4)2^08 in this connection, so that his ex-
periments were plainly made imder diflFerent conditions from
those proposed.

2d. Determination of Co. I prefer to determine €Jo as neutral
€oS04. We have the advantage here of determining the
■Go in the form of a compound of higher atomic weight, thus
lessening our errors of analysis. The cobalt is precipitated from
the solution of the double sulphates as €oS, wasned with hot sul-
phid of ammonium water, dned and roasted and then converted
into sulphate with the aid of aqua regia and sulphuric acid. The
results are good, although not so exact as in the case of the fiisi-
ble double sulphates. I also determined €o as metal by H after
its precipitation as hvdrated oxyd by 01 and acetate of soda,
a method proposed by Popp.* Instead of 01 I found it far
more convenient in practice to use a strong solution of Ol^Oi,
which is a powerful oxydizing agent, and as it can be kept,
becomes a useful laboratory reagent This method, however,
does not always give good results, on account of the difficulty of
washing the hydrated ^o%^% till free from alkali

8d. Determination of the N. This was done, first by the
modification of Bunsen'sf method proposed by Dr. W. Gibbs,^:
which space will not permit me to give in detail The reader
is therefore referred to the original paper. Suffice it to say, that
for nitrates and nitrites to which it is applicable, it is an excel-
lent method. The loss of the combustiou-tube can be nothing
but N and HjO. I^ therefore, the €a01a tube retains the HjO
accurately, the remainder is N. Whether, therefore, the H2O is
accurate, or high from hygroscopic moisture, the nitrogen must
be accurate, being subject to no souh^e of error.

The nitrogen was also determined by volume, using the Spren-
gel-pump. This method was first proposed by Frankland§ for
nitrogen in analyses of potable waters, but he did not apply it
then to organic analysis. The application, however, had hieen
made in this laboratory by Dr. Gibos some months before Dr.
Frankland's paper was received. A vacuum is first made in the
combustion-tube by means of the mercury-pump. The delivery-
tube from the pump is then connected with a Simpson's receiver
and the vacuum destroyed by heating some "GOj placfed in the
anteriorportion of the tube. Any excess of €0^ is absorbed
by the KHO solution in the receiver. The combustion is now
made. A final vacuum is then obtained by the pump and the
receiver disconnected. After a short time the gas is transferred
to a eudiometer-tube and measured according to the methods of
gas-analysis. The results are generally good, but depend upon

* Ann. Oh. u. Ph., 131, 363. + Ann. Ch. u. Ph., 72, 40.

X This Jofir., xxxvii, 360. § Jour. Ch. Soc., 1868, p. 90.

Digitized by VjOOQ IC

0. C. Marsh on Oretaceotis and Tertiary Birds. 205

the vacuum obtained at the beginning and end, on the perfect
combustion, and on the perfect transfer of the gas.

In conclusion I would present my grateful acknowledgments
to Dr. W. Gibbs, to whose kindness I am indebted for the selec-

Online LibraryRodolfo Amedeo LancianiThe American journal of science and arts → online text (page 24 of 109)