which, however, is limited, varying with the amount of base metals present; should they be present in considerable quantities the residue must be weighed, after drying at 1 1 0 ’ C., t o avoid an error caused b y their oxidation. Ordinarily. hon-ever, this oxidation is so small t h a t i t need not be taken into account. I n the separation of carbonates and ignition of residue, the loss will also include. small amounts of water, volatile and organic matter. The residue left insoluN r a H.v.iPsrx I R E ble in acetic acid contains all the silica originally presAGRICULTURAL E X P E R I 3 I E N T S T A T I O S , ent in the ore. The silica is now determined b y DCRHAX.NE\\. I I A > l P S € i I R E . volatilizing the same as silicon fluoride. I h e presence of sulfides of iron, zinc and lead must be considered, and for this reason I have found i t best t o VALUATION OF FLUORSPAR. oxidize t h e sulfides in the residue b y heating with merB y E. BIDTEL. curic oxide before evaporating with hydrofluoric Received January 2 , 1912. acid. In this operation lead sulfide is entirely oxitlI n the commercial analysis of fluorspar the deter- ized t o lead sulfate, b u t the ferruginous zinc sulfide minations usually required are calcium fluoride, is b u t slightly acted upon and remains unaffected b y silica and calcium Carbonate; in some particular cases hydrofluoric acid, in so far as final results are conlead, iron, zinc and sulfur. We have received several cerned, as shown b y analysis of synthetical samples calls for percentage of barium sulfate b u t t h e most of known percentages. careful qualitative tests have failed t o shox- the 111 regard t o t h e determination of calcium fluoride, presence of this mineral. as \Till be shown later, in I avoid decomposing t h e sample and eliminate the either n-hat is known as Rosiclare or Fairview fluor- calcium fluoride b y treating the residue obtained from spar. the silica determination with hydrofluoric acid t o There is no practical method for the separation transform the iron oxide into iron fluoride and e s of fluorine from silica and of determining fluorine tracting the same together with the lead and zinc direct in its ore, which answers the commercial re- b y solution of ammonium acetate containing ammoquirements of a mine laboratory. The usual prac- nium citrate. The method as used at present follows: tice follon-ed b y most chemists is to determine the Weigh into a small Erlenmeyer flask one gram of total calcium oxide and t o calculate from this the t h e finely powdered sample, add I O cc. of ten per amount of calcium fluoride, after deducting the cent. acetic acid, cover with a short-stemmed glass amount equal t o the calcium carbonate present, this funnel and heat on a water bath for one hour, agitalatter being determined from a separate sample in ting from time t o time. Filter through a j cm. ashthe usual wa>- b y absorbing the liberated carbonic less filter, mash nTith warm v a t e r four times and burn acid in soda-lime tubes. This method would be correct off the filter paper in a weighed platinum crucible if all the calcium present were combined with either at a temperature as low as possible. The loss in weight carbonic acid or fluorine and all t h e carbonic acid minus O . O O I ~gram (the amount of calcium fluoride present combined with calcium. The first assumption soluble in acetic acid under t h e conditions named) is probably correct as all our experiments go t o prove is calcium carbonate. this, b u t the latter we know is wrong. for lead carbonAdd t o the residue in the platinum crucible about ate, zinc carbonate and iron carbonate are frequently one gram of yellow mercuric oxide in t h e form of an Con- emulsion in water; break up any hard lumps t h a t present, especially in the gravel fluorspar. sequently, the calculated amount for calcium carbonate m a y have formed; evaporate t o dryness and heat n-ill be too high and the calcium fluoride in consequence t o a dull red h e a t ; cool and weigh. Add about 2 cc. will be too lev,. of hydrofluoric acid and evaporate to dryness, repeatI n T-iv7.v of t h e above error. I have found it better ing this operation twice, using I cc. of hydrofluoric t o diss,.jl\.e all carbonates from the sample b y acetic acid in each of the last two operations. Add a few acid. lea\-ing the calcium fluoride and silica as a resi- drops of hydrofluoric acid and some macerated fildue (UII the filter read!* for the determination of both ter paper, as recommended b y Dittrich, then a few silica ant1 calcium fluoride. I t must he remembered drops of ammonium hydroxide to precipitate the iron t h a t calcium fluoride is slightly soluble i n acetic acid. and evaporate t o dryness. Heat t o a dull red heat, .I have reduced this solubilitJ- t o a factor which I . cool and weigh; the loss in v-eight is reported as silica. use in m?- calculatians. Generallh- the amount of KOK add 2 cc. uf hydrofluoric acid and a few drops of calcium carbonate present is greatlj- in excess oi the nitric acid, cover the crucible with its lid and place other c n i - b ( n i a t e s . so t h a t frequently i t will be sufi- on a moderately warm water bath thirty minutes ; cientl!- accurate fur commercial purposes t o report remove the lid and evaporate t o dryness. If the conthe amount soluble in acetic acid after deducting t h e tents of the crucible are not now perfectly white, amount of calcium fluoride soluble in this acid as evaporate again with hydrofluoric acid, add a few calcium carbonate. I t is to be remembered t h a t a t drops of hydrofluoric acid anti I O cc. of the solution this point there is oliportunit>- for another small error of ammonium acetate (this ammonium acetate is as given in column 4 gives t h e soluble arsenic oxid not combined with lead oxid. The method outlined in this paper gives a rapid and accurate procedure for determining soluble arsenic oxid in commercial lead arsenate, It is pointed out t h a t the solubilit-\- of lead arsenate affects the total amount of soluble arsenic oxid and a correction is worked out which gives the amount of soluble arsenic oxid not combined with lead oxid.
T H E JOCR,VAL OF I n ; D C S T R I A L A S D E.VGIiYEERISG CHEJIISTRY.
202
prepared as follows: neutralize very carefully 400 cc. of eighty per cent. acetic acid in a liter flask with strong ammonium hydrate; add twenty grams of citric
Average samples of different car loads showed the following percentage, one gram of material being taken for analysis:
CaSOl CaFl Average plus obtained b y W-eighed decomposition equal 0.29 per as CaF2. cent. with H?SOa. to CaS04. 1
2
3
4
5
6
Keystone ground.
., . ..
A - l Ground
Gravel.. . . . . . . . .
I II :;::: I 99.07 99.01 99.06 96.81 96.77 96.86
89.38 86.76 Gravel.. . . . . . . . 86.87 86.85 82.14 Gravel. . . . . . . . , . 8 2 . 1 3 182.08 66.02
.
Unwashed s a n d . . 66.04
172.75 172.65 172.63 168.80 168.73 168 89 155.72 155.69 155.71 151 .OO 151.24 151.27 143.03 143.00 143.10 114.80 114.86 114.82
99.06 99.01 98.99 96.76 96.80
1
0.42 1.47 1.42 1 .45
86.72 66.75 82'03 82 .OO 82.06 65.83 65.87 65,84
I 1
1
3 . 3 31 3.29 4.93 0 4.95 82.32
66.14
acid and fill up t o the mark with strong ammonium hydrate). Digest on a boiling mater bath for thirty minutes ; filter a n d wash b y decantation with hot water containing a small amount of the ammonium acetate solution just described, then with pure, hot water; ignite in the same crucible and weigh. The residue should be perfectly white and should be pure calcium fluoride. T o test this, a d d 2 cc. of sulfuric acid, heat to decompose the calcium fluoride, and evaporate the excess of sulfuric acid ; repeat this operation, using I cc. of sulfuric acid. Now weigh the calcium sulfate, a d d five or six grams sodium carbonate, fuse and dissolve in hydrochloric acid in slight excess; should a white precipitate occur a t this point i t will indicate the presence of barium. To prove this method I selected clear crystals of fluorspar, pulverized the same and purified it by treating with acetic acid, hydrofluoric acid and ammonium acetate; the washed, dried and ignited powder was decomposed by sulfuric acid and the calcium sulfate weighed. By taking an average of six determinations 0.387j gram of this fluorspar formed 0.67jj5 gram of calcium sulfate equal to 9 9 . 9 7 per cent. calcium fluoride. One gram of this fluorspar treated as described showed the following results as a n average of twelve determinations: Gram. Loss in weight by treating with acetic acid. . . . , . . . . . . . . . . . Loss in weight by treating with HgO and H F . . , . . , . . . . . . . . Loss in weight by treating with ammonium acetate. . . , . . . . .
0.0015 0.0002 0.0012
The amount of calcium fluoride lost in these different manipulations was 0 . 0 0 2 9 gram, or 0 . 2 9 per cent., if we take one gram for analysis. I now analyzed mixtures of material of known percentage similar in composition to fluorspar a n d obtained the following results, using one gram for analysis:
.
Taken., , , , . . .
.
Taken., . . , . . . Taken., , . , , , , . Found .... , , . . .
CaF2.
SiOz.
CaC03.
PbS.
85.11 85.0i 80.05 80.03 78.45 78.41
6.66 6.61 7.71 7.70 8.32 8.29
5.59
1.50
1.14
ZnFeS.
2.50
1.82
1.40
1.43
5.57 7.92 7.87 10.40 10.34
0.45
99.31
96.85
1
So1ub 1e in acetic acid less 0.15 h\-erage. per cent. Average.
SiO?.
t
89.28 89.29
>Iar., 1912
f
{
I t 1 t
I
1.45
(I:;
0.88
(I:;:
1
1
1 1
6 .' 0 18 9 6.12
$
1 5 .' 5 2 7 15.65
5
t
8.6i
13.75 13.i6
1
13.77
I
)
I
1
CaC03 equal tocoz
0.24
0.116
0.26
0.88
0.35
0 80
4.23
1.75
3.98
4.68
1.95
4 43
6.13
2.61
5.98
15.58
6.83
15 52
1
4.93
I
13.80
0.20
4.27 4.17 4.25 4.72 4.69 4.63
3 31
8.71
8.5 70 9
{ i(
con.
I
1
The calcium sulfate obtained by decomposition of fluorspar was fused with sodium carbonate and the cake dissolved in hydrochloric acid. The solution was perfectly clear, showing a total decomposition of the fluorspar and the absence of barium. As our fluorspar does not contain any gypsum and heavy spar, I did not investigate its influence on the calcium fluoride determination, b u t it is evident t h a t small amounts of calcium sulfate will be leached out, if present, together with the carbonates. Barium sulfate is easy to detect and can be determined as usual. The only constituent of injurious influence would be any silicate containing calcium. F. Julius Fohs, in Bulletin 9 of the Kentucky Geological Survey, states the following silicates occurring associated with Kentucky fluorspar: Apophyllite, Datolite. Desmine, Epidote, Laumontite and Natrolite. Apophyllite and Natrolite cannot be present in the fluorspar because there are no alkalies to detect. The same with Datolite. I could not find any trace of boron; Desmine and Laumontite cannot be present in any considerable amount, there being only veq- small amounts of aluminum in the spar. I have made a support of aluminum which enables me to handle six crucibles a t a time and I can .finish six analyses in ten hours. LABORATORY FAIRVIEW FLUORSPARASD LEADCo., GOLCONDA. ILL.
DETERMINATION O F MANGANESE I N STEEL. BY
JAMES
J. BOYLE.
Received December 15, 1911.
After successively trying for some time the color method, Johnson's lead peroxide-sodium arsenite titration method and the ammonium persulphate-sodium arsenite titration method for manganese in steel with varj-ing success, the writer has adopted a suggested
