62 
PROCEEDINGS OP THE AMERICAN ACADEMY 
VI. 
RESEARCHES ON TIIE INORGANIC ACIDS. 
By Wolcott Gums, M.I)., 
Hum ford Professor in Harvard University 
(Continued from Vol. XVI. p. 139.) 
Presented May 24th, 
PHOSPHO-MOLYBDATES. 
Tiie application of molylxlic oxide to the separation and estimation of 
phosphoric acid has given a special interest to the phospho-molybdates, 
and they have accordingly been studied more or less completely by 
several chemists. The most thorough investigations which we possess 
are those of Debray,* Rammelsberg,t and Finkener,J but particular 
salts have been examined by others, and these will be noticed under 
the appropriate special headings. 
Phospho-molybdates appear to be formed whenever phosphoric acid 
or a soluble phosphate is brought into solution with a molybdate, the 
presence of a free acid not being essential. They are also formed 
when phosphates and molybdates are fused together, when molyb- 
dates insoluble in water are dissolved in phosphoric acid, when 
molybdic oxide is digested with an alkaline phosphate, and when in- 
soluble phosphates and molybdates are treated together with a dilute 
acid. As a class, they are better defined and more easy to obtain 
pure than the phospho-tungstates which in many respects they closely 
resemble. When phospho-molybdates of fixed alkaline bases are 
heated, they at first give off water of crystallization, and by careful 
heating may be obtained anhydrous. In some cases, however, molyb- 
dic oxide is volatilized even from salts containing fixed alkaline bases. 
* Bull. Soc. Cliim., [2.] v. 404. 
t Berichte der Deutsclien Client. Gesellschaft, Zelinter Jahrgang, p. 1776. 
J Ibid., Elfter Jahrgang, p. 1038. OF ARTS AND SCIENCES. 
63 
I did not succeed in obtaining well-defined pyro-phospho-molybdates or 
pyro-phospho-tungstates, though of course the residues of the ignition 
of the acid salts may be regarded as such. When a phospho-molyb- 
date is dissolved in ammonia-water and a current of sulphydric acid 
gas is passed into the hot solution, sulpho-molybdates are formed in 
large quantity. This reaction distinguishes the phospho-molybdates 
from the phospho-tungstates which are not decomposed rnider the 
same circumstances. 
Analytical Methods. — The determination of the sum of the per- 
centages of molybdic and phosphoric oxides was usually effected, as in 
the case of the phospho-tungstates, by precipitating the two oxides 
together by mercurous nitrate with addition of mercuric oxide to neu- 
tralize the free nitric acid. It is best to precipitate from a boiling 
solution, and to boil for a short time after adding mercuric oxide. 
This last must always be in small excess. On account of the volatility 
of molybdic teroxide, it is not possible to determine directly the sum 
of the weights of the two oxides by simple ignition, but the difficulty 
may be readily overcome by the following process. The filter with 
the mercurous salts is to be cautiously heated in a platinum crucible 
properly inclined to the vertical axis of the flame until the filter is 
completely carbonized. On then regulating the heat and the supply 
of air, the carbon may be readily burned off, leaving a mass of mer- 
curous salts mixed with more or less mercuric oxide, no weighable 
amount of molybdic teroxide being lost. An accurately weighed 
quantity of anhydrous normal sodic tungstate in fine powder is then 
to be added, and the contents of the crucible carefully mixed together 
with a stout platinum wire previously weighed with the crucible 
itself. The whole is to be heated at first by radiation from a small 
iron dish, and afterward directly, until a clear white fused mass is 
obtained. A second ignition and second weighing will determine 
whether every trace of mercury has been expelled. It is almost 
needless to remark, that all these operations must be conducted under 
a flue with a good draught. This process gives excellent results, and 
is much less tedious than would perhaps be supposed. 
After the estimation of the phosphoric oxide the molybdic teroxide 
is best determined by difference from the sum of the weights of the 
two oxides found as above. No really good general method for the 
quantitative separation and estimation of molybdic oxide has yet been 
given, at least no one which is sufficiently accurate to serve as a check 
upon the method above described. The ammonium salts of this series 
are most simply analyzed by igniting them directly with sodic tung- 64 
PROCEEDINGS OP THE AMERICAN ACADEMY 
state, when the loss of weight corresponds to the sum of the water 
and ammonia. 
As in the case of the phospho-tungstates, the quantitative determi- 
nation of phosphoric oxide is a matter of considerable difficulty. The 
method of separation by means of magnesia mixture has been carefully 
studied by Dr. Gooch, to whose paper I have already referred.* Dr. 
Gooch found it necessary to precipitate the ammonio-magnesic phos- 
phate a second time, a single precipitation giving an error amounting 
sometimes to G or 7 % of the phosphoric acid present. After re-solu- 
tion and precipitation by ammonia, the mean error amounted to only 
O.G5%, which makes an almost insensible correction when the quan- 
tity of phosphoric oxide is small. In a few instances I have applied 
this correction after a double precipitation, but I prefer to employ the 
following method, which gives an almost perfect separation from 
molybdic teroxide. The phosphoric oxide is first precipitated from a 
hot solution as ammonio-magnesic phosphate, the supernatant liquid 
after complete subsidence carefully decanted upon an asbestos filter, 
the precipitate washed with magnesia mixture and ammonia, then 
redissolved in the least possible quantity of hot dilute chlorhydric acid 
and reprecipitated with ammonia. After complete subsidence ami 
decantation, the precipitate is boiled with successive portions of a 
solution of ammonic sulphide. A more or less dark orange-red so- 
lution of ammonic sulpho-molybdate is always obtained at first, but 
after two or three repetitions of the process the ammonic sulphide 
added remains colorless on heating. The ammonio-mugnesian phos- 
phate is then filtered upon the asbestos filter already employed. In 
place of this method I have sometimes employed the following modifi- 
cation, which gives, I think, equally good results. After the first 
precipitation the phosphate is to be redissolved, and tbe bot solution 
precipitated at once by ammonic sulphide in excess. The precipitated 
phosphate is then to be boiled two or three times with ammonic 
sulphide as above. Whatever inaccuracy is inherent in this method 
depends, in my judgment, upon the fact that, as Dr. Gooch has 
shown, the determination of phosphoric acid by means of magnesia 
is, under the most favorable circumstances, a less accurate process 
than has been supposed. 
The determination of ammonia and the alkalies was effected by the 
methods already described in the case of the phospho-tungstates. 
Water must be estimated by ignition with sodic tungstate, as there is 
* Proceedings of American Academy, Vol. XV. p. 5o. OF ARTS AND SCIENCES. 
65 
often volatilization of molybdic teroxide when a phospho-molybdate is 
ignited at a temperature sufficient to expel its water. The analyses 
require great care and no small amount of practice to insure good 
results. As in the case of the phospho-tungstates, the alkaline bases 
are best determined by difference. 
Twenty-four Atom Series. — Phospho-molybdic acid. The acid of 
this series was first obtained by Debray, who prepared it by boiling 
ammonic phospho-molybdate with nitro-muriatic acid, and allowing 
the solution to evaporate spontaneously. I find that this is a good 
method of obtaining the acid, but the following details should be 
observed. The bright yellow ammonic phospho-molybdate should be 
first dried, and then heated with a large excess of strong aqua-regia 
in a casserole over an iron capsule to serve as a radiator. In this 
manner the decomposition proceeds very regularly and without suc- 
cussions. When it becomes necessary to add fresh acid, the super- 
natant liquid should be allowed to settle completely and then be poured 
off carefully. Fresh acid may then be added, and the process, which 
is at best a slow one, continued. When the ammonium salt has disap- 
peared, the liquid is to be evaporated until the excess of nitric and 
chlorhydric acids has been expelled. On standing, large bright yellow 
octahedral crystals are obtained from the very concentrated solution. 
These may be redissolved and recrystallized, but there is always 
some loss in the process of purification, because solution in water pro- 
duces more or less decomposition of the acid with formation of a pale 
greenish white crystalline body. This substance passes very readily 
through a filter, and the solution of the acid must be allowed to settle 
completely before the clear supernatant liquid is brought upon the fil- 
ter. Debray obtained three different hydrates of phospho-molybdic 
acid, to which he gave, respectively, the formulas 
20 Mo08 . P205.3 H20 + 21 aq, 
20 Mo03. P205.3 H20 -f 48 aq, . 
and 
20 Mo03. P205.3 II20 -f 38 aq. 
Unfortunately he has not given either the methods or the com- 
plete results of his analyses. In the first hydrate he found 13.30%, 
in the second 23.40%, and in the the third 19.60% of water. 
I obtained the acid only in transparent octahedral crystals which 
had a bright yellow color. Of these crystals, dried by pressure with 
woollen paper, 66 
PROCEEDINGS OF THK AMERICAN ACADEMY 
0.9945 gr. lost by ignition witli W04Na2 0.2362 gr. = 23.75% water. 
1.45*8 gr. gave 0.0713 gr. P2O.Mg2 = 3.12% P206. 
The analysis leads to the formula 
24 Mo03. P206.3 II20 + 59 aq, 
which requires: — 
24 Mo08 
345G 
Calo’d. 
73.31 
Found. 
73.13 
P,06 
142 
3.01 
3.12 
C2 II20 
1116 
23.68 
23.75 
4714 
100.00 
The phosphoric oxide was determined by double precipitation and 
treatment with ammonic sulphide. The molybdic oxide was estimated 
by difference. The crystallized acid effloresces so readily that the pre- 
cise determination of the water is difficult. In a portion of the crystals 
which had effloresced in a very marked degree, — 
0.9873 gr. lost on ignition with W04Na3 0.1760 gr. = 17.82% water 
2.2472 gr. gave 0.1163 gr. P207Mg2 = 3.31 % P205 
The ratio of the molybdic to the phosphoric oxide is in this analysis 
also 24:1; and, if we compute the results of both analyses for an an- 
hydrous compound of the two oxides, we find : — 
24 Mo08 
3456 
Calc’d. 
96.06 
95.91 
95.97 
p,o. 
142 
3.94 
4.09 
4.03 
3598 
100.00 
100.00 
100.00 
The analyses leave, I think, no reasonable doubt as to the ratio of 
the two oxides. Phospho-molybdic acid therefore corresponds in com- 
position with phospho-tungstic acid, the ratio of the two oxides being 
24:1, as given by Finkener,* and not 20:1, as stated by Debray. 
With respect, however, to the number of atoms of water in the crys- 
tallized octahedral hydrate, I may remark that, while the analysis 
agrees best with the formula given, 
24 Mo03. P206.3 II20 -f 59 aq, 
* Loc. cit. OP ARTS AND SCIENCES. 
67 
it is much more probable that the acid really contains an atom less of 
water, and that its formula, apart from the question of basicity, is 
24 Mo03 . P205.6 H20 -f- 55 aq, 
like 
24 W03 . P20. . 6 H20.+ 55 aq, 
already described. This formula requires 23.38% water, instead of 
23.75%, as found. Debray found 23.40%. As already stated, the 
crystals analyzed were dried by pressure with woollen paper, after 
draining off a syrupy mother liquor, and may therefore not have been 
perfectly free from extraneous water. Finally, the analyses of Finke- 
ner led also to the formula with G1 atoms of water, and I shall adopt 
this as the definite constitution of the octahedral hydrate. Finkener’s 
work has not yet been published in detail; but from the abstract which 
he has given, it clearly appears that we owe to him the establishment 
of the true constitution of the only phospho-molybdic acid yet obtained. 
As already mentioned, there are two other hydrates of phosplio-tung- 
stic acid, having, respectively, the formulas 
24 W03. P205.6 II20 + 47 aq, 
and 
24 W03 . P205.6 H20 + 34 aq. 
The two hydrates of phospho-molybdic acid described by Debray 
would correspond to the formulas 
24 Mo03 . P205.6 II20 -f- 24 aq, 
and 
24 Mo03 . P205.6 II20 -j- 43 aq, 
if we suppose them, as is most probable, to belong to the 24-atom se- 
ries. The first formula requires 13.05%, the second 19.66% water; 
Debray found 13.09 % and 19.60 %. Finkener obtained still another 
hydrate, containing about 32 atoms of water, basic water included. 
Phospho-molybdic acid dissolves very readily in water, forming a 
colorless liquid which has a strong acid reaction. As already stated, 
the solution is always accompanied by a slight decomposition, with for- 
mation of a very pale greenish white crystalline substance. A pre- 
cisely similar decomposition is observed in the solution of the corre- 
sponding phospho-tungstic acid. The crystals lose all their water when 
slightly ignited. According to Finkener, three atoms of water remain 
at 140° C. The solution readily expels carbonic dioxide from the 
alkaline carbonates. The question of the basicity of the acid will be 
discussed farther on. 68 
PROCEEDINGS OF TI1E AMERICAN ACADEMY 
24 : 3 Ammonic Pliospho-molybdate. — The constitution of the beau- 
tiful yellow salt which is formed when an excess of a mineral acid is 
added to a solution containing molybdic and phosphoric oxides and a 
salt of ammonium, has long been in dispute. The analyses of Svan- 
berg and Struve,* Nutzinger,f Sonnenschein,t Lipowitz,§ and Selig- 
solm,|| gave results which differed very sensibly from each other, 
according to the method of analysis employed. Debray gave the 
formula 
20 Mo03 . P203.3 (NH4)20 -f 3 aq, 
but without the details of his analysis. More recently the subject has 
been examined with great care by Finkener,1f who has arrived at the 
conclusion that, though the percentages of water and ammonia may vary 
within wide limits, the ratio of the molybdic and phosphoric oxides is 
always as 24 : 1. 
With respect to the preparation and properties of the yellow ammo- 
nium salt, I have little to add to what has been done by these chemists. 
I repeatedly prepared the salt for analysis, usually by mixing solutions 
of ammonic molybdate — 7:3 salt — and phosphate, adding nitric acid 
in excess to the solution, and boiling. When the mixed solution is 
boiled for a short time, the precipitation of the yellow salt is complete 
after standing until the liquid becomes cold. In the publication of this 
result, which is important in analysis, I have been anticipated by Atter- 
berg; ** but I propose in another paper to give the results of my work 
on the quantitative determination of phosphoric acid, and will then give 
ample details. 
As regards the composition of the yellow phospho-molybdates of 
ammonium, my #esults do not agree with those of Finkener, as I think 
I have evidence that, as in the case of the phospho-tungstates, there 
are series of phospho-molybdates in which the ratio of the molybdic to 
the phosphoric oxide is as 20 :1, as 22 :1, and as 24:1. In one prepa- 
ration, — 
1.1492 gr. lost on ignition with W04Na2 0.0827 gr. NH3 and H20 
= 7.20% 
* Journal fur prakt. Chemie, xliv. 291. 
t Pharmaceut. Vierteljahresschrift, iv. 649. 
1 Journal fiir prakt. Chemie, liii. 342. 
§ PoggendorfTs Annalcn, cix. 135. 
|| Journal fiir prakt. Chemie, lxvii. 470. 
1T Loc cit. 
** Benchte der Cliem. Gesellschaft, 1881, p. 1217. 69 
OP ARTS AND SCIENCES. 
0.5905 gr. lost on ignition with W04Na2 0.0432 gr. NII3 and II20 
= 7.31% 
1.7158 gr. gave 0.1027 gr. P207Mg2 = 3.83% P205 
0.9806 gr. “ 0.0567 gr. P20-Mg2 = 3.70% P205 
1.8903 gr. « 0.1321 gr. NH4C1 =3.20% (NH4)20 
In these analyses, the first determination of the phosphoric oxide 
was made by double precipitation only, without subsequent treatment 
with ammonic sulphide; but in the second, this reagent was employed1 
in the manner above described. The ratio of Mo03 to P205 is almost 
precisely 24: 1, and the analyses correspond closely with the formula* 
24 Mo03. P205.3 (NH4)20 -f 24 Mo03. P206.2 (NII4)20. II.,0 
-|- 16 aq, 
which requires: — 
Calc’d. 
Mean. 
48 Mo03 
6912 
89.05 
89.00 
2 PA 
284 
3.66 
3.75 
3.70 
3.83 
5 (NH4)20 
260 
3.35 
3.39 
3.39 
17 H20 
306 
3.94 
3.86 
3.81 
3.92 
7762 
100.00 
100.00 
Acid salts of similar type occur frequently in the class of phospho- 
molybdates, as in that of phospho-tungstates. 
24:1 Croceo-cobalt Salt. — The disposition of the cobaltamines to 
form highly crystalline compounds, together with their well-defined and 
various degrees of basicity, led me to study the relations of these bases 
to the phospho-molybdic acids. This had already been done to a cer- 
tain extent with the 5:1 atom series by results I 
shall cite in connection with that series. Neither roseo-cobalt nor 
luteo-cobalt forms well-defined salts with 24 : 1 phospho-molybdic acid. 
I had therefore recourse to croceo-cobalt,* the oxide of which may be 
written 
Co2(NH8)8(N02)40, or, briefly, CcO. 
The chloride of this series gives no precipitate with solutions of 7 :3 
amnionic molybdate, or of hydro-disodic phosphate; but in an acid solu- 
tion of these two salts a solution of the chloride throws down a beau- 
tiful bright yellow highly crystalline salt, which may be washed with 
cold water. The portion analyzed was dried on woollen paper only. 
Of this salt, — 
* Proceedings of American Academy, Vol. X. p. 1. 70 
PROCEEDINGS OF THE AMERICAN ACADEMY 
1.0728 gr. gave 0.8133 gr. MoO.. + P20. = 75.81% 
i.4520 gr. “ 0.4719 gr. P205 ° = 2.96% 
This corresponds to 72.85% MoOa by difference, and 24.19% of 
CcO and water by the loss. The analyses agree very closely with 
the formula 
24 Mo03 . P203. CcO . 2 HaO + 21 aq, 
which requires : — 
24 MoOs 
3456 
Calc’d. 
72.82 
72.86 
PA 
142 
2.99 
2.96 
CcO 
734 
15.47 ) 
f 
23 II20 
• 414 
8.72 } 24,1<J 
| 24.19 
4746 
100.00 
Under the microscope this salt is seen to consist of fine yellow felted 
needles. It is very slightly soluble in cold water, but is soluble in a 
rather large quantity of boiling water, giving an orange-yellow solution, 
with a strongly acid reaction. The solution gives with argentic nitrate 
a very insoluble sulphur-yellow tlocky precipitate, which after a time 
becomes crystalline, and a pale yellow fiocky precipitate with mercu- 
rous nitrate. No precipitate is formed with cupric sulphate or baric 
chloride. The salt could not be recrystallized ; it is interesting as a 
particularly well-defined soluble acid salt of the 24: 1 atom series. 
24:2 Acid Potassium Salt. — This salt was prepared by boiling 
together solutions of potassic molybdate and phosphate, adding an ex- 
cess of nitric acid, and boiling the whole for some time. As in the 
case of the ammonium salts, the precipitation is greatly facilitated by 
this process, taking place very slowly in the cold. The salt obtained 
was in very minute crystals, bright yellow, and but slightly soluble in 
cold water. Of this salt, — 
0.7772 gr. lost on ignition 0.0128 gr. = 1.64% water 
0.7962 gr. “ “ 0.0130 gr. = 1*66% “ 
1.1703 gr. gave 1.0895 gr. Mo03 + P203 = 93.10% 
1.3263 gr. “ 0.0779 gr. P20-Mg2 " = 3.76% P,0, 
1.3033 gr. “ 0.0778 gr. “ = 3.82% “ 
The phosphoric oxide was twice precipitated as ammonio-magnesic 
phosphate. The analyses correspond with the formula 
24 Mo08 . P,05.2 K.,0 . II20 + 3 aq, 
which requires : — 71 
OF ARTS AND SCIENCES. 
Calc’d. 
24 Mo03 
3456 
89.55 
89.31 
p2o5 
142 
3.69 
3.76 
3.82 
2K20 
189 
4.90 
5.25 
4H20 
72 
1.86 
1.66 
1.64 
3859 
100.00 
Twenty-two Atom Series. — In the paper already referred to, Ram- 
melsberg has described several salts in which he found the ratio of 
molybdic to phosphoric oxide as 22: 1. Unfortunately, he has not 
given the method of analysis which he employed, and in a question of 
so much difficulty and delicacy it is, to say the least, extremely desira- 
ble to know what degree of precision may be expected in the analyses. 
As his results appear to be supported by my own, I shall adopt them, 
leaving to the further progress of analytical chemistry the final settle- 
ment of the few doubtful points. 
22:3 Ammonium Salt. — Rammelsberg found for the neutral salt 
of this series the formula 
22 Mo03. P206.3 (NH4)20 -f 12 aq, 
which corresponds, except as regards the amount of water of crystalli- 
zation, with a phospho-tungstate which I have already described, — 
In one preparation of a yellow insoluble ammonium salt exactly 
resembling the corresponding salt of the 24-atom series, — 
1.6885 gr., lost on ignition with W04Na2 0.0873 gr. = 5.17% NII3 
and H20 
1.7764 gr. gave 0.6200 gr. P207Mg2 = 4.17% P2Os. 
1.9024 gr. “ 0.6029 gr. “ = 4.01% “ 
1.2334 gr. “ 0.6262 gr. NH4C1 = 4.23% (NH4)20. 
The salt was dried for some time in pleno over sulphuric acid, and 
had evidently lost water of crystallization. If we deduct the remain- 
ing water, 0.94%, and calculate the analysis for an anhydrous salt, 
we have for the formula 
22 W03. P205.3 (NH4)20 —J— 21 aq. 
22 Mo03. P205.3 (NH4)20; 
22 MoOg 
31G8 
Calc’d. 
91.41 
91.68 
PA 
142 
4.09 
4.05 
3(NH4)20 
156 
4.50 
4.27 
3466 
100.00 
, 72 
PROCEEDINGS OF THE AMERICAN ACADEMY 
In another preparation,— 
1.0324 gr. lost on ignition with 0.0922 gr. = 8.93% 'Nllg 
and II20 
2.0670 gr. gave 0.1255 gr. P20TMg2 = 3.88% P205 
2.0352 gr. “ 0.1220 gr. “ = 3.84% “ 
These analyses lead to the formula 
22 Mo03 . P203.3 (NH4)20 + 9 aq, 
which requires: — 
22 Mo08 
3168 
Calc’d. 
87.32 
87.21 
PA 
142 
3.91 
3.88 3.84 
3 (NII4)tO 
156 
4.291 
- 8.77 8.93 
9 11,0 
162 
4.48 ) 
3628 
100.00 
If from the analyses of the two salts above described we calculate 
the composition of the combination of molybdic and phosphoric oxides 
supposed to be isolated, and compare this with the percentages calcu- 
lated upon the two hypotheses of a ratio of 22 :1 and a ratio of 24:1, 
we have: — 
22 MoO, 
3108 
Calc’d. 
95.76 
I. 
95.76 
II. 
95.76 
Calc’d. 
96.06 
3456 
24 Mo03 
142 
4.24 
4.24 
4.24 
3.94 
142 
i3A 
3310 
100.00 
100.00 
100.00 
100.00 
3598 
In botli eases the phosphoric oxide was precipitated twice, but the 
ammonia-magnesian phosphate was not treated with ammonic sulphide. 
According to the results of Dr. Gooch already cited, the probable 
error of this method does not exceed 1 % in excess of the quantity of 
phosphoric oxide present. It appears, therefore, that the correction 
to be applied to the phosphoric oxide in the above analyses does not, 
at most, exceed 0.04%. The mean of Dr. Gooch’s analyses would 
require a deduction of 0.02% only. The yellow ammonium salt 
analyzed by Rammelsberg corresponds to the formula 
22 Mo08 . V.205.3 (NH4),0 -f 12 aq, 
which requires (Rammelsberg): — 
22 MoO, 
3168 
Calc’d. 
86.04 
86.45 
PA 
142 
3.86 
3.90 
3(NH4)sO 
156 
4.24 
3.25 
12 11,0 
216 
5.86 
5.77 
3682 
100.00 
99.37 73 
OF ARTS AND SCIENCES. 
Rammelsberg gives these figures as the means of several analyses 
which agree well with each other, but it must be admitted that a 
closer correspondence with the percentages required by the formulas 
would have been desirable. The comparison is not given in his 
paper. The air-dried salt loses all its water over sulphuric acid. The 
three atoms of basic water, if we assume their existence, must there- 
fore be united by a very feeble affinity. Rammelsberg has also 
analyzed the corresponding potassic salt of the same series. I here 
give his results, for the sake of comparison with the formula : — 
22 Mo03 
3168 
Calc’d. 
83.17 
84.43 
p2o5 
142 
3.73 
3.78 
8 K20 
283 
7.43 
6.86 
12 H20 
216 
5.67 
5.55 
3809 
100.00 
100.62 
This salt loses all its water between 120° and 140°. In judging 
the results of these analyses, as well as of those which I have given, 
it must be carefully borne in mind that the salts themselves cannot be 
recrystallized, and that consequently their absolute purity cannot be 
guaranteed. Moreover, if — as I believe I have shown — there are 
very similar salts which represent three series in which the ratios of 
the molybdic and phosphoric oxides are respectively as 24:1, 22 :1, 
and 20: 1, we may, at least occasionally, have mixtures of the salts of 
three, or of any two series. The difficulty here is precisely that which 
occurs in the case of the phospho-tungstates. 
44: 2 Acid Potassium Salt. — This salt was prepared by boiling 
a mixture of potassic molybdate and phosphate with nitric acid in 
excess, when a beautiful yellow crystalline powder separated. This 
was washed with cold water and dried on woollen paper. Of this 
salt, — 
0.9850 gr. lost on ignition 0.0521 gr. = 5.28oj0 water. 
0.8983 gr. gave 0.7943 gr. MoOa -j- P205 = 88.42 
2.0617 gr. gave 0.1201 gr. P207Mg2 = 3..72% 
These analyses lead to the formula 
44 Mo08.2 P205.5 K20 . II20 + 21 aq, 
or 
22 MoOg. P205.3 K20 + 22 Mo03. P205.2 K20 . H20 + 21 aq, 
which requires: — 74 
PROCEEDINGS OF THE AMERICAN ACADEMY 
44 Mo03 
633 G 
Calc’d. 
84.62 
84.70 
2 P205 
284 
3.79 
3.72 
5 K20 
472 
6.30 
6.30 (diff.) 
22 HaO 
396 
5.29 
5.28 
7488 
100.00 
The salt is therefore the acid 6alt corresponding to a neutral salt 
with the formula 
22 Mo03 . P,0.5.3 Iv20. 
Rammelsberg’s analyses agree better with the formula of the acid 
salt given above than with that of the neutral compound assumed by 
him. 
Twenty Atom Series. — The only salt of this series which I have 
obtained is one of ammonium prepared like the salts already described, 
having like these a fine yellow color and a very fine-grained crystalline 
structure, and like them hut slightly soluble in water. Of this salt, — 
1.093G gr. lost on ignition with W04Na2 0.0729 gr. = G.66% NIIa 
and H20 
1.8183 gr. lost on ignition with W04Na2 0.1155 gr. = 6.35% NH3 
and H20 
0.8862 gr. gave 0.G153 gr. NII4C1 =4.12% (NH4)20 
1.3213 gr. gave 0.6224 gr. P,07Mg2 = 4.19% P.206 
1.5135 gr. gave 0.G349 gr. P207Mg2 = 4.31% P205 
The salt was dried on a water-bath, and afterward over sulphuric 
acid. The phosphoric oxide was precipitated twice, but not treated 
with amnionic sulphide. The analyses lead to the formula 
GO Mo03.3 P205.8 (NII4)30.11,0 + 11 aq. 
which requires: — 
no Mo03 
8640 
Calc’d. 
89.09 ) 
- 93.48 
89,21 1 93.52 
3 Pa05 
426 
4.39 i 
4.31 4.19) 
8 (NH,),0 
416 
4.29 ) 
■ 6.52 
4,12 l 6.50 
12 11,0 
216 
2.23 j 
• 2.54 \ 
9698 
100.00 
If we calculate the composition of the mixed oxides of molybdenum 
and phosphorus existing in this salt we have: — OF AliTS AND SCIENCES. 
75 
Calc’d. 
20 Mo03 
2880 
95.30 
95.39 
P203 
142 
4.70 
4.61 
3022 
100.00 
100.00 
It will be seen that the ratio is here very nearly as 20 : 1. This may 
however be merely accidental, and farther researches are necessary to 
fully establish the existence of a 20-atom series. 
According to Debray a solution of argentic nitrate gives with one of 
phospho-molybdic acid a precipitate which soon becomes crystalline, 
and which has the formula 
20 MoOs. P205.7 Ag20 + 24 aq. 
Such a salt would possess a twofold interest, first, as another evi- 
dence of the existence of a 20-atom series of phospho-molybdates, and, 
secondly, as showing that the acid of the series may unite with more 
than six atoms of base. On mixing the two solutions as above, I ob- 
tained a precipitate in small indistinct crystals of a greenish yellow color. 
These crystals were soluble in hot water, but the solution was quickly 
decomposed with precipitation of a white powder. Under the micro- 
scope with a high power and transmitted light the salt appeared to 
consist of small tabular crystals mixed with a few long yellow prisms 
of very different habitus. Of this compound, — 
1.3604 gr. lost by ignition with W04Na2 0.0692 gr. water = 5.08% 
2.1099 gr. gave 0.8287 gr. AgCl = 31.63% Ag20 
0.6733 gr. gave 0.2619 gr. AgCl = 31.44% Ag20 
2.1099 gr. gave 0.0928 gr. P20-Mg2 = 2-81% P205 
The phosphoric oxide was determined in the filtrate from the ar- 
gentic chloride by double precipitation and treatment with amnionic 
sulphide. The ratio of the molybdic to the phosphoric oxide is as 
21 :1, but the formula which most nearly represents the analysis is 
22 Mo03. P205. 7 Ag20 -J- 14 aq, 
which requires, — 
22 Mo08 
3168 
Calc’d. 
61.08 
60.57 
p2o5 
142 
2.74 
2.81 
7 Ag20 
1624 
31.32 
31.44 31.63 
14 1I20 
252 
4.86 
5.08 
5186 
100.00 76 
PROCEEDINGS OF TIIE AMERICAN ACADEMY 
The only conclusion which can fairly be drawn from the analysis is 
that there is at least one phospho-molybdate in which the number of 
atoms of b:ise exceeds three. It is certain that the salt does not rep- 
resent a perfectly definite and homogeneous compound, and it may 
possibly be a mixture of the 20-atom salt, 20 Mo03. P2Cb • 6 Ag20, 
and an acid molybdate of silver, 2 MoOs. Ag20, nearly in atomic pro- 
portions. lly dissolving the salt in nitric acid and evaporating, Debray 
obtained another salt in small brilliant yellow crystals. For this salt 
he proposes the formula 
20 M0O3 . P205 • 2 Ag20 + 7 aq, 
but as usual he has given no analyses. 
Eighteen Atom Series. — I have myself met with no salts belonging 
to this series, but according to Finkener* there ai’e sodium salts corre- 
sponding to the general formula 
18 Mo03 • I*A (3 — x) NaaO (25 -f- x) aq- 
These salts are yellow and easily soluble. 
Sixteen Atom Series. — 16:3 Ammonium Salt. — In preparing the 
5 :3 atom ammonium salt a white crystalline precipitate was formed, 
insoluble in cold, but soluble with decomposition in much boiling water, 
and easily soluble in ammonia. In this salt dried over sulphuric 
acid, — 
0.5100 gr. lost by ignition with 0.0722 gr. = 14.16% NH3 
and II20 
1.1653 gr. gave 0.1259 gr. NH4C1 = 5.25% (NII4)20 
0.8114 gr. gave 0.0658 gr. P207Mg2 = 5.19% 1,206 
The analysis corresponds with the formula 
16 Mo03. P205.3 (NII4)20 + 14 aq, 
which requires,— 
Calc’d. 
Found. 
1G Mo03 
2304 
80.73 
80.65 
p„o. 
142 
4.97 
5.19 
3 (NH4)aO 
156 
5.46 
5.25 
14 II20 
252 
2854 
8.84 
100.00 
8.91 
* Loc. cit., p. 1639. OF ARTS AND SCIENCES. 
77 
Five fo One Series. — Salts of this series were discovered at an 
early period in the history of the subject by Zenker.* The ammonium 
salt was analyzed by Zenker f and Werncke,J and recently by Ram- 
melsberg. § Debray obtained the same salt, but has published no 
analyses. Rammelsberg also obtained the corresponding potassium 
salt, as well as an acid salt of the same series. The alkaline salts are 
colorless, and separate in well-defined crystals, which are usually easily 
soluble in water. The acid of the series, as Debray has stated, can- 
not be obtained by the decomposition of its salts, being resolved by 
acids into free phosphoric acid and salts of the 24-atom series. The 
decomposition may probably be expressed by the equation 
24 (5 Mo03 . PA • 3 H20) = 
5 (24 MoOs. P205.3 H,0) + 19 (P2Os. 3 H,0). 
All the neutral salts are tribasic (old style) or more correctly hexa- 
tomic, but well-defined acid salts exist in which the ratio of the molyb- 
dic oxide to the fixed base is as 10:5. Such salts have been obtained 
by Rammelsberg and by myself. The salts of the higher series are de- 
composed by alkalies, as stated by Debray, salts of the 5-atom series 
and alkaline molybdates being formed. Conversely, when a mineral 
acid is added to a solution of an alkaline salt of the 5-atom series, a 
salt of a higher series is formed, frequently as a yellow crystalline 
precipitate. The neutral salts of this series hitherto described have 
respectively the formulas 
5 MoO,. PA.3 K20 —|— 7 aq. 
5 Mo03. PA • 3 (NH4)P + 7 aq. 
5 MoOs. PA • 3 Na2° 4- 14 aq. 
5 MoOs. PA • 3 AS-2° + 7 aq- 
5 : 3 Phospho-molybdate of Ammonium. — This beautiful salt ap- 
pears, as already stated, to have been first obtained by Zenker. It 
is readily obtained by dissolving together five molecules of amnionic 
molybdate and two of ammonic phosphate, and evaporating the solu- 
lution, when beautiful prismatic crystals, with a glassy lustre, separate. 
These may easily be purified by recrystallization. The salt is readily 
soluble in hot, less easily in cold water. The solution has an acid 
reaction. Zenker’s analyses, as well as those of Werncke, agree 
closely with the formula 
5 Mo03 . P.205.3 (NH4)20 + 7 aq, 
* Journal fur prakt. Cheinie, lviii. 256. 
$ Zeitschrift fur Analyt. Cheinie, xiv. 12. 
t Loc. cit. 
§ Loc. cit. 78 
PROCEEDINGS OF THE AMERICAN ACADEMY 
in which formula the pliospho-molybdic acid is regarded as tribasic. 
Debray gives the same formula, without details of analysis, and 
Ilammelsberg has very recently again analyzed the salt, confirming 
the results of Zenker. The salt in question is particularly interest- 
ing, first, because the number of atoms of molybdic oxide is uneven ; 
and secondly, because the basicity of the acid appears to be 3, and not 
6, even when the salt has separated from neutral solutions. 
Jorgensen * has described two well-defined crystalline salts belonging 
to this series and having according to his notation respectively the 
formulas 
Co2 (NHs)10C12 . (5 Mo08.2 P04II) 
and 
Co2(NII3)10C12 . (5 MoOa. 2 P04NII4). 
I should write these 
and 
5 Mo03 . P205. Co2(NH3)10C1 A • Hs0 
5 MoOs. P,05. Co2(NHJ10C12O2(NH4)2O. 
It will readily be seen that both salts correspond to the acid repre- 
sented by the formula 
5 MoOg. P205.3 II20. 
Acid 10 : 5 Ammonium Salt. — When ammonic phosphate is dis- 
solved in boiling water, and molybdic oxide is added in small portions 
at a time, the oxide readily dissolves, but a greater or less quantity 
of a white insoluble crystalline salt is formed. The liltrate deposits on 
evaporation large colorless crystals, which appear to be either trimetric 
or monoclinic. Of these crystals,— 
1.112G gr. lost on ignition with W04Na2 0.2076 gr. = 18.66% NII3 
and II20. 
1.2062 gr. “ “ “ 0.2425 gr. = 18.71% “ 
1.2165 gr. “ “ “ 0.2247 gr. = 18.47% “ 
0.9263 gr. gave 0.1912 gr. P807Mg2 = 13.20% P2Os 
1.0540 gr. “ 0.2196 gr. “ = 13.32% “ 
1.1824 gr. “ 0.3018 gr. NH4C1 = 12.41% (NH4)20 
1.0183 gr. “ 0.2563 gr. “ = 12.23% “ 
1.6430 gr. “ 0.4168 gr. “ =12.32% “ 
* Journal fur prukt. Chemie,’[2.] xviii. 209. OP ARTS AND SCIENCES. 
79 
The analyses lead to the formula 
10 Mo08.2 P206.5 (NH4)20 . II20 + 6 aq, 
or 
5 Mo03 . P205.3 (NII4)20 + 5 Mo03. P205.2 (NII4)20. H20 + G aq, 
which requires: — 
10 MoOg 
1440 
68.24 
68.13 
• • • 
. . . 
. .. 
2 PA 
284 
13.46 
13.26 
13.20 
13.32 
• . • 
5 (NHJ.O 
260 
12.32 
12.32 
12.41 
12.23 
12.32 
7 H20 
126 
2110 
5.98 
100.00 
6.29 
6.15 
6.34 
6.39 
The phosphoric oxide was determined by double precipitation only, 
without subsequent treatment with amnionic sulphide. The percent- 
age is a little lower than that required by the formula, which is 
unusual; but the general agreement of the analyses with the formula 
is satisfactory. Rammelsberg has described an acid potassium salt 
with the formula 
10 Mo03 . P205.5 K,0 . II20 -f 19 aq. 
It is therefore at least probable that we shall find another ammonia 
salt with 20 atoms of water, and another potassium salt with 7 atoms. 
Zenker has described another potassic salt to which he gives the 
formula, — as I should write it, — 
9 Mo08 . P205. 4 K20.2 H20 -f 18 aq; 
but the results of his analyses differ very widely from the percentages 
required by the formula, and on repeating his process I obtained 
only the 10:5 atom salt of Rammelsberg. The formula given above 
for this salt requires 11.11% P205. I found 11.22%. 
Rammelsberg * has also described a white insoluble potassium salt 
to which he gives the formula 15 Mo03. P205.5 K20, but without any 
statement of his analyses. 
ARSENIO-MOLYBDATES. 
Compounds of arsenic and molybdic oxides have been described 
by Seyberth f and by Dei)ray4 Seyberth obtained an acid with the 
formula, — as I should write it, — 
* Loc. cit. 
t Berichte der Cliem. Gesellscliaft, 1874, p. 391. 
t Comptes Rendus, lxxviii. 1408. 80 
PROCEEDINGS OF THE AMERICAN ACADEMY 
7 MoOs As205.3 II20 + 11 aq, 
and the three corresponding salts : — 
7 MoOg. As205 . (NH4)20.2 II20 + 4 H20 
7 Mo03 . AsjOg. 3 BaO 
7 ]\IoOg. As2Og. 3 Ag-O. 
Debray obtained the acids and one or two salts of two different 
series, which may be represented respectively by the formulas: — 
20 Mo03 . As./).. 3 ir20 -J- 24 aq 
20 MoO,. AggOj. 3 K,0 
20 MoOg. As205.3 (NH4)aO 
G Mo03. As206.3 H.,0 + 13 aq 
6 MoO,. As205.4 (NII4)20 + aq 
6 Mo08 . As2Os . (NII4)20 + 4 aq 
6 Mo03. As205 . Na20 —}— 12 aq. 
Debray considers the formula of the 20-atom ammonium salt as 
probable only, and regards the water determination in the corre- 
sponding acid as not quite certain. Neither Seyberth nor Debray has 
described the analytical methods employed, or given the details of 
the analyses. 
I have found it most advantageous to separate arsenic from mo- 
lybdic oxide by precipitating with magnesia mixture, redissolving the 
ammonio-magnesic arsenate, and precipitating a second time with 
ammonia after adding a little magnesia mixture. The ammonio-mag- 
uesic arsenate may be digested with ammonic sulphide without de- 
composition ; but after the second precipitation it does not retain 
molybdic oxide, and the subsequent treatment is therefore unneces- 
sary. To determine the sum of the molybdic and arsenic oxides 
I precipitate the two together with mercurous nitrate and mercuric 
oxide, in the manner already described for the estimation of molybdic 
and phosphoric oxides, filter upon paper, and after drying roll up the 
filter and its contents and ignite cautiously in a porcelain crucible. 
By slow and careful heating the filter may be completely burned with- 
out loss of molybdic or arsenic oxides, this result being attained by 
the oxygen of the mercurous and mercuric oxides present. A 
weighed quantity of sodic tungstate is then to be added in fine powder, 
the mass well mixed in the crucible, and then cautiously heated until 
mercury is completely expelled, and after cooling a white fused mass 
remains. A second or even a third heating is necessary to insure a OP ARTS AND SCIENCES. 
81 
perfectly constant weight. The difference between the percentage 
of arsenic oxide, As206, and the sum of the percentages of the arsenic 
and molybdic oxides, gives the percentage of molybdic oxide with 
a very fair degree of approximation. In these salts the water must 
always be determined by ignition with sodic tungstate or some similar 
compound, since both arsenic and molybdic oxides are volatile. 
Sixteen to one Series. — When solutions of ammonic arsenate and 
acid molybdate (7 : 3 salt) are mixed, a beautiful white crystalline pre- 
cipitate is thrown down, which is very insoluble in cold water but 
dissolves slightly in boiling water, giving, however, a turbid solution. 
The salt is readily soluble in ammonia water. The portion analyzed 
was well washed on a filter with cold water and dried on woollen 
paper. In this salt, — 
1.1322 gr. lost on ignition with W04Na2 0.1G36 gr. NIIS and II20 
= 14.45% 
1.3389 gr. gave 0.2481 gr. NII4C1 = 9.00% 
1.4276 gr. “ 0.1478 gr. As207Mg2 = 7.68% As205 
The analyses lead to the formula 
16 Mo03. As205.5 (NII4)20 . tl20 + 8 aq. 
which requires: — 
16 MoO„ 
2304 
Calc’d. 
77.94) or 
77*97 Us 65 
7.68 I85'65 
As206 
230 
7.78 }t,S’72 
5 (NH4)20 
260 
S-]l 1 14.28 
9-°° 1 14.45 
9 H20 
162 
5.49 / 
5.45 J 
2956 
The salt may have lost a little ammonia in drying. When potassic 
arsenate and acid molybdate are mixed, a similar salt is formed. A 
solution of arsenic acid gives at once in solutions of acid amnionic or 
potassic molybdate a beautiful white crystalline precipitate, insoluble in 
cold water, but soluble in a large quantity of boiling water, forming 
cloudy solutions which pass freely through a filter. These may serve 
as starting-points for new investigations. The arsenio-molybdate above 
described is not perceptibly altered by long boiling with nitric acid, 
but the existence of higher compounds containing 22 or 24 molecules 
of molybdic to one of arsenic oxide appears at least extremely probable. 
The phospho-molybdates and arsenio-molybdates now known with 
some degree of certainty are as follows : — 82 
PBOCEKDINGS OF THE AMERICAN ACADEMY 
24 Mo08 . P206.6 II20 -f 47 aq Mo24P2071(HO)12 -f 47 aq 
21 MoO, . PjO, . 0 II20 + 413 aq Moa4P2On(HO)12 + 43 aq 
24 MoOg . P206.6 II20 -f 24 aq Mo24P2On(HO)12 -f 24 aq 
24 MoOg . P206.2 K,0.4 II20 Mo24P2071(KO)4(HO)8 
24 MoOg . P208 . CcO . 6 H20 4 18 aq Mo24P2O71(CcO2)(HO)10 + 18 aq 
48 MoOg . 2 P206.5 (NH4)20.1I20 -f 10 aq Mo48P4O148(N1I4O)i0(HO)2 +10 aq 
22 MoOg . P206.3 (NII4)20.3 II2() + 9 aq Mo22P208,(NH40)6(HO)8 + 9 aq 
22 MoOg . P206.8 (NH4)20.3 H20 4- 0 aq 4 (i aq 
22 MoOg . P206.3 (NH4)20 Mo22P2068(NII40)6 
22 MoOg . 1’206.3 K.,0.8 II20 + 9 aq Mo2,P2066(KO)„( 1I0)6 + 9 aq 
22 MoOg . P206.7 Ag,0 + 14 aq AgO)14 4- 14 aq 
44 MoOg . 2 P206.6 K20 . H20 4 21 aq Mo44P4O186(KO)10(HO)2 + 21 aq 
GO MoOg . 3 P206.8 (N1I4)20 . II204-11 aq Mo60P6O186(NH4O)16(IIO)2 + 11 aq 
18 MoOg . P205 . Na20.5 HaO + m aq Mo18P2O68(NaO)2(IIO)10 m aq 
18 MoOg . P206.2 Na,0.4 II20 + n aq Mo]8P2058(NaO)4(IIO)8 4 n aq 
16 MoOg . P206.3 (NII4)20.3II20 4 11 aq Mo16P2047(NII40)6(IIO)6 4- 11 aq 
5 MoOg . P206.3 Na20.3 H20 -f 11 aq Mo6P2014(NaO)6(HO)6 + 11 aq 
6 MoOg . P206.3 (NH4)20.3 H20 4- 4 aq Mo6P.,014(NH40)6(I10)6 -f 4 aq 
5 MoOg . P206.3 K.,0.3 H20 -f 4 aq Mo6P2014(KO)6(HO)6 -f 4 aq 
6 MoOg. P206.3 Ag20.3 H20 -f 4 aq Mo6P2014(AgO)6(IIO)6 + 4 aq 
10 MoOg . 2 Pa06.6 K,0.11,0 4- 19 aq Mo10P4O84(KO)10(HO)2 4- 19 aq 
10 MoOg . 2 1>206.6 (NII4)26.1I20 + 0 aq Mo10P4()84(NII4O)10(Ilb)2 + 6 aq 
20 MoOg . As206.6 HaO 21 aq Mo20As2O69(IIO)12 4 21 aq 
20 MoOg . As206.3 K20 Mo20As2O62(KO)6 
20 MoOg . As206.3 (NH4)20 Mo20As2O62(N1I4O)6 
16 MoOg . As206.5 (NII4)20 . IL20 + 8 aq MoI6As2O47(NII4O)10(IIO)2 -f 8 aq 
7 MoOg . As206.6 H20 -f 8 aq Mo7As2O20(HO)12 4 8 aq 
7 MoOg . As206 . (NH4)20.5 II20 Mo7A82O20(NH4O)2(IIO)10 
7 MoOg . As206.3 BaO Mo7As202g( Ba02)8 
7 MoOg . A8206.3 Ag20 Mo7As2028( AgO)e 
6 MoOg. As206.0 II20 4 10 a0 Mo6A82017(IIO)12 4 10 aq 
6 MoOg. As206.4 (NII4)20 4 a(l Mo6As2Ol9(NII40)8 
6 MoOg . As206 . (N1I4)20.2 H20 4 2 aq Mo6As2O.20(NH4O)2(IIO)4 4 2 aq 
6 MoOg . As206 . Na./> . 5 1I20 4 7 aq Mo6As2O17(Na())2(IIO)10 4 7 aq 
For the convenience of comparison with the corresponding com- 
pounds of tungsten, I have in writing these formulas as far as possible 
assumed that all the phospho-molybdic and arsenio-molybdic acids con- 
tain 12 atoms of hydroxyl, or, in the language appropriate to the old 
notation, are six-basic. With the material before us, we are now pre- 
pared to discuss the question of the basicity of the phospho-tungstates OF ARTS AND SCIENCES. 
83 
and phospho-molybdates as well as of the corresponding arsenic com- 
pounds. 
The general results to which the study of the phospho-molybdates 
and arsenio-molybdates has led are as follows : — 
1. The phospho-molybdates form a series of which the lowest term 
contains five atoms of molybdic to one of phosphoric oxide, and the 
highest twenty-four atoms of the former to one of the latter. 
2. As in the case of the phospho-tungstates, the greater number of 
the molybdenum compounds contain an even number of atoms of tung- 
stic oxide. The homologizing term is therefore 2 Mo03 for these 
cases. 
3. By far the greater number of phospho-molybdates contain three 
atoms of fixed base (old style), or, in more modern language, may be 
considered as derived from acids containing six atoms of hydroxyl. 
Anhydrous compounds of this type occur, and are not always simply 
residues obtained by heating salts which may be considered as acid, as 
containing, for example, 3 R20.3 H20. It seems therefore necessary 
to admit the existence of acids of the general type 
n Mo03. P205.3 H20, 
which may, however, stand in the relation of pyro-acids to other acids 
of the type 
n Mo08. P205.6 H20. 
4. On the other hand, while no single phospho-molybdate containing 
more than three atoms of fixed base for one of phosphoric oxide has 
been obtained in a state of indubitable purity, it is probable that there 
is at least one salt with six or more atoms of fixed base. I refer to 
the silver salt which I have expressed by the formula 
22 Mo08. P20.. 7 Ag20 + 14 aq. 
5. Setting aside the evidence derived from the analogy of the phos- 
pho-molybdates and phospho-tungstates, there is at present no sufficient 
proof of the existence of a series of phospho-molybdates or arseuio- 
molybdates containing more than three atoms of fixed base. Such 
purely negative evidence must not be too highly regarded. 
6. As in the case of the phospho-tungstates, there exists a class of 
phospho-molybdates in which the ratio of the number of atoms of base 
to that of the number of atoms of phosphoric oxide is as 5 : 2, the num- 
ber of atoms of molybdic oxide being even. 
Since the publication of my work on the phospho-tungstates and 84 
PROCEEDINGS OF THE AMERICAN ACADEMY 
arsenio-tungstates a paper by Sprenger* on the phospho-tungstates has 
appeared. Sprenger has examined, with a single exception, only the 
compounds of the 24 : 1 series, and has added a number of new salts, 
which, so far as regards their constitution, fully confirm my own results. 
The compounds described, belonging to the 24-atom series, are the 
following: — 
24 WO, . PA . 3 11,0 + 58 aq 
24 WO,. PA • 3 BaO + 58 aq 
24 WO, . PA • 2 BaO . H.,0 + 58 aq 
24 WO,. PA • BaO . 2 H*0 + 58 aq 
24 WO,. PA • 8 Cu20 + 58 aq 
24 WO,. PA • 3 Ag,0 + 58 aq 
24 WO,. PA • AgaO . 11,0 + 58 aq. 
Sprenger’s formula for the octahedral acid agrees with that which I 
had given if we consider the acid as tribasic. The other salts which 
he has described are new, and form a valuable addition to our knowl- 
edge of this class of compounds. It is well worthy of notice, that in all 
of his salts, the acid included, the number of atoms of water is the 
same. The acid with 58 atoms of water of crystallization forms, there- 
fore, a complete and stable molecular structure in which 2, 4, or G atoms 
of hydrogen are replaceable. I do not recall any other series in which 
this constancy of crystalline water occurs, at least to the same extent. 
Sprenger has also obtained a salt of the 22-atom series which is of 
much interest. This is the barium salt 
22 WOa. P205.7 BaO + 59| aq, 
and its special interest depends upon the fact, first, that the ratio of the 
tungstic to the phosphoric oxide is as 22:1, and, secondly, that the 
salt contains seven atoms of fixed base, or, in other words, must be con- 
sidered as derived from an acid containing at least fourteen atoms of 
hydroxyl. Sprenger asserts that he has obtained the corresponding 
acid, and it is to be hoped that he will pursue the subject farther. 
This barium compound furnishes additional evidence of the independent 
existence of a series in which the ratio is 22 : 1, and in addition it 
renders more probable the formula which I have given for Debray’s 
silver salt, 
22 Mo03. P205.7 AgaO + 14 aq. 
From these two tolerably well-established cases it would appear that 
* Journal fiir prakt. Chemie, xxii. 418. 85 
OF ARTS AND SCIENCES. 
we are not justified in holding that the phospho-tungstates, phospho- 
molybdates, and corresponding arsenic compounds, have a basicity of 
which the higher limit is six. I may here mention that I shall here- 
after describe a vanadio-molybdate of ammonium the analyses of which 
agree well with the formula 
18 Mo03 . V205.8 (NII4)20 —|— 15 aq. 
The risk of drawing hasty conclusions from purely negative evi- 
dence is particularly great in discussing the degree of basicity of this 
whole class of compounds, but 1 shall endeavor to show that it is possi- 
ble to devise structural formulas which will embrace and explain all 
degrees of basicity which appear to be possible under the general con- 
ditions of the problem. 
We may, as in the case of the alkaline tungstates already discussed, 
assume that both tungsten and molybdenum are hexatomic, and, as in 
that case, we may start from the commonly received formula for po- 
tassic dichromate, 
Cr02 - O - K 
I 
o 
I 
Cr02 - 0 - K 
which may be equally well applied to hexatomic tungsten, 
o 
II 
o - W - 0 - K 
i 
0 
I 
O = W - O - K 
II 
0 
If we further suppose that the separate terms of the structural 
formulas are symmetrically arranged, and take a 6:1 phospho-tung- 
state 
6 W03. P205 • 6 H20 or 6 W02. PA • (HO) J2 
as an illustration, we may, with at least a certain degree of proba- 
bility, express the structure as follows: — 86 
PROCEEDINGS OF THE AMERICAN ACADEMY 
HO - WOa = WO,-OH 
i i 
IIO-WO, — W02-0II 
I I 
HO-WOa — WO,-OH 
I I 
o o 
I I 
3 (IIO) = PO-O-OP = (OII)3 
Thi* formula explains the basicity of the acid satisfactorily. It 
also shows that, as six atoms of hydroxyl are united with phosphorus, 
and six with tungstic oxide, there should be theoretically a limiting 
case corresponding to an acid containing six atoms of hydroxyl, and 
represented by the formula 
6 W03. P,05.3 II20 or G W02.03. P205. (HO)6 
and structurally by 
woa = wo2 
i \ 0/ i 
wo, -o- wo, 
II II 
wo, -O- wo, 
I 1 
o o 
I I 
3 (HO) = PO - O - OP = (OIT)a 
According to this view six atoms of hydroxyl are always associated 
with phosphorus, or, as the case may be, with arsenic. I consider this 
view of the subject by far the more probable. At tbe same time, how- 
ever, it is also possible that we may have the structural formula, 
HO-WO, = W02-0II 
i i 
HO - WO, — WO,-OH 
i i 
II0-W02 — WO, - OH 
i i 
0 0 
I / o \ ' 
op — PO 
in which all the atoms of hydroxyl are associated directly with tung- 
sten, and in the present state of our knowledge we can only decide the OF ARTS AND SCIENCES. 
87 
question upon general grounds of probability, so that our conclusions 
are at best uncertain. Finally, both formulas being at least possible, 
it may be that there are two isomeric modifications of each series of 
acids represented respectively by the formulas above given. There is 
no present evidence of the existence of such isomeric modifications in 
the case of phospho-tungstates, phospho-molybdates, or the correspond- 
ing arsenic series; but Marignac has shown that there are two isomeric 
series of silico-tungstates, which he calls respectively silico-tungstates 
and tungsto-silicates, and it may be that the difference between these 
depends upon differences in the mode of combination, precisely similar 
to those which I have pointed out above. I shall return to this sub- 
ject in the general discussion of my results. With respect to the two 
linking terms, 
II II 
0 0 0 0 
i i i / 0 N i 
3 (HO) = PO - 0 - OP = (OH), and OP — PO 
no assumption is made which is not in perfect accordance with com- 
monly accepted views of the subject. 
We may now consider the most general case, that, namely, in which 
there are twenty-four atoms of tungstic or molybdic, to one of phos- 
phoric or arsenic oxide. We have for an acid of this type 
24 W03. P205.6 H20 or 24 W02.018. P205. (HO)12 
and in accordance with the principles above laid down the structural 
formula may be written: — 88 
PROCEEDINGS OF THE AMERICAN ACADEMY 
wo, = wo,, 
I I 
o o 
I I 
wo2 = wo2 
I I 
o o 
I I 
wo2 = wo2 
I I 
o o 
I I 
wo2 = wo2 
I I 
o o 
I I 
wo2 = wo2 
I I 
o o 
I I 
wo2 = wo2 
I I 
o o 
I I 
wo2 = wo2 
I I 
o o 
I I 
wo2 = wo2 
r i 
o o 
I I 
wo2 = wo2 
I I 
o o 
I I 
IIO - W02 — W02 - OH 
I I 
HO-WO, — W02-0H 
I I 
II0-W02 — W03 - OH 
I I 
O O 
I I 
3 (HO) = PO-O-OPe (OII)3 OP ARTS AND SCIENCES. 
89 
The case of an acid containing for twenty-four atoms of tungstic 
oxide six atoms of hydroxyl may easily be deduced from the above, 
upon the principle explained in the first example cited. Without again 
writing the cumbrous formula, it may easily be seen that the cases 
of acids containing more than twelve atoms of hydroxyl, if such really 
exist, are embraced in the above-given structural formula, and that in 
such cases there will be two variations in the mode of combination of the 
hydroxyl, similar to the two which occur when there are six or twelve 
atoms of hydroxyl. The structural formula given would explain sim- 
ply and naturally the tribasic character of all known phospho-molyb- 
dates and phospho-tungstates containing twenty-four atoms of metallic 
oxide, since in these all the hydroxyl may be united with phosphorus 
exclusively, or with tungsten exclusively. It only remains to consider 
the case of the compounds having for one atom of phosphoric or arsenic 
oxide an uneven number of atoms of metallic oxide, as, for instance, 
the 5 :1 and 7 :1 series. In these cases also there exists, as has been 
shown, a second and derived series, of which the successive terms are to 
be regarded as formed from those of the first series by doubling the mo- 
lecular weight and dropping an atom of fixed base. Thus, we have 
5 Mo03. P206.3 H20 and 10 Mo03.2 P205.5 K20 . H20 + 19 aq 
7 Mo03 • As2°5 • 3 H2° 14 WOs • 2 P2°5 • 3 Na20 . H20 + 42 aq 
22 Mo03 • p2°5 • 3 H2(4 44 m°03 • 2 P206.5 K20 . H20 -j- 21 aq 
24 Mo03 • P2°5 ■ 3 H2° 48 Mo°3 • 2 P2°o • 5(NII4)20. H20+16 aq 
All these salts appear to have an acid reaction. They may all be 
regarded as acid six-basic salts, and it is easy to see that the two series 
may be reduced to one by doubling the formulas of all the terms on the 
left, so that we shall have a single series, of which the successive terms 
are 
10 M0O3 • 2 P2°5 • 6 IT2° 
12 Mo03.2 As20. . 6 H20 
14 W03.2 P20. . 6 H20 
48 Mo03.2 P205.6 H20 
This view in no wise excludes acids or salts of a higher degree of 
basicity. It has the advantage of bringing all the compounds to- 
gether, and of being more completely in accordance with what we 
know of the constitution of salts belonging to simpler types. The 
structural formulas which I have given — provisionally, of course — 90 
PROCEEDINGS OF THE AMERICAN ACADEMY 
may easily be modified to suit this view, and will all be symmetrical, 
and suggestive of various possible isomerisms. 
The study of other complex inorganic acids will, doubtless, throw 
further light upon the subject, and to it I shall continue to devote my 
leisure. It already begins to appear that inorganic compounds may 
possess an unexpected degree of complexity, and that very wide fields 
of research in inorganic chemistry are still open. 
( To be continued.)