5
T H E J O U R N A L OF I N D L - S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
766
crystallization takes place or t h e mass will spatter. After cooling, take u p with about 2 j cc. water, shake until dissolved, a n d again evaporate just t o fuming in order t o remove the last traces of lower oxides of nitrogen. Allow t o cool, takk u p with j o cc. cold water (the stannic sulfate dissolves more quickly b y using less water at first) and when all is in solution, a d d z j cc. more water, I O cc. of 40 per cent K I , and titrate immediately with t h e N / I O thiosulfate. If the alloy contains a small amount of lead, some lead sulfate remains undissolved, and t h e solution a t the e n d point is yellowish (owing t o t h e lead iodide) instead of the creamy white of cuprous iodide, b u t the presence of lead has n o effect upon t h e titration. a t least in moderate amounts A copper-tin alloy containing about I O per cent of tin a n d 8 per cent lead analyzed as follows: N O
1.. . .. . . . ..... 2.. , , .., ..... 3.. , , . . , , , , ., . . 4., . . . . .., . . . .. 5,....,,.......
Sample Gram 0.3127 0.3227 0.3303 0,3108 0.3327
Cc. NazS%Os 39.49 40.79 41.70 39.21 42.02
-4verage Cu FOUND Per cent Gram Per cent Cu 0.2474(8) 7 9 . 1 4 ; 0.2556(3) 7 9 . 2 2 79.16 0.2613 (4) 79 12,: 0.2457 (4) 0.2633(4) i9'11
I:?;
I n the analyses indicated as 4 and j above, the alloy was decomposed with nitric, t h e metastannic acid filtered off, a n d t h e copper determined in t h e filtrate after evaporation with sulfuric. Too much stress cannot be laid upon the fact t h a t a second fuming of t h e solution is necessary: 0 . 0 5 gram of metallic tin, treated with nitric, fumed once with sulfuric, a n d diluted t o 7 j cc., gave a blue color with K I a n d starch requiring 0 . jj cc. t o discharge it. When this same amount of tin was fumed twice, potassium iodide and starch gave no color in the diluted solution. As regards t h e time necessary, determinations I , z and 3 , cited above, were r u n in one hour. T h e filtering, washing and evaporating necessary when t h e ordinary method was used, made t h e time about five hours. ( I n this, no correction was made for a n y copper remaining with the metastannic acid.) The method proposed is simple, accurate. and rapid; the a d r a n t a g e s of carrying out t h e analysis in t h e same container are obvious. I n conclusion. it must be remembered t h a t no metals which separate iodine from potassium iodide m a y be present, such as antimonic or ferric salts; t h e method has naturally t h e limitations of t h e iodide method in this respect. It is simply the desire of the writer t o show a n extended use of this excellent method, t h u s making i t available under apparently adverse conditions. 614
EAST1 l i T H
STREET,
CLEVELAND,
OHIO
TlTRATION OF NITRATES WITH FERROUS SULFATE B y FREDC. BOWMAN A N D \\-. W. SCOTT Received M a y 29, 1915
T h e great variety of methods of determining nitrates found in chemical literature shows t h e need felt b y chemists for simpler ways of making this important analysis. Most methods now in use involve distillation or evaporation or the'use of special appara-
V O ~7. , NO. 9
tus, all of which t a k e a good deal of time and attention. The method proposed b y the authors is as easy and quick as a n y ordinary titration, has a , w i d e range of usefulness and is accurate enough for most purposes. BIBLIOGRAPHY
The first mention of t h e method was made b y Grossart1 in 1847. He titrated nitrates in boiling 60 per cent sulfuric acid with a ferrous sulfate solution, using ferricyanide as an outside indicator. He gives no details or test analyses. K O doubt his results were less accurate t h a n he thought. Mohr2 announced t h e discovery of t h e method in 1861. He dissolved t h e sample in sulfuric acid (diluted I t o 9) and titrated a t 70" t o 80" C. with a ferrous sulfate solution containing zoo grams of t h e salt per liter, using as t h e end point t h e appearance of a brownish color. He declared t h a t it was a good technical method, b u t Fresenius3 and E d e r , 4 who examined it later, condemned it strongly. I n 1899 van Deventerj carried out the reaction over mercury in a Crum tube with exclusion of air, using fairly strong sulfuric acid as a medium and a weak ferrous sulfate solution as reagent. T h e process was awkward t o handle. S o comment on these methods was found in the journals, a n d evidently they failed of acceptance. T h e y were all based on t h e reduction of t h e nitrate t o K O and not t o N203, as in t h e method given below, for estimations in presence of sulfuric acid. N o mention of acids other t h a n sulfuric was found in t h e literature. ESTIU.4TIOX
O F SITRIC ACID IX P R E S E K C E O F A R S E N I C ACID
The writers devised t h e method given below, originally, for t h e purpose of estimating nitric acid in arsenic acid. T h e original form of t h e method was as follows: Dilute t h e sample t o be tested t o I O O cc. with nitrefree arsenic acid of such strength t h a t there is not more t h a n 20 per cent water in the mixture. Then titrate with a ferrous sulfate solution, containing 264.7 grams FeSO4.7HsO and jo cc. strong sulfuric acid per liter ( I cc. = 0 . 0 2 g. "03). Standardize t h e solution on a known amount of nitric acid. The following reaction occurs: 6FeS04 zHN03 3HzSO4 = 2NO 3Fe2(S04)3 4HsO The ferrous sulfate forms a dark brown color, which disappears on stirring as long as a n y nitric acid remains. T h e end point is t h e first permanent brown and is readable t o 0 . 0 3 cc. During the first t w o thirds of t h e titration, t h e ferrous sulfate is taken u p very quickly; then a lively evolution of N O commences a n d continues t o the end. It would be thought t h a t t h e oxygen of t h e air might reoxidize the N O
+
1
+
+
"Sur un nouveau dosage de l'acide azotique e t des azotates," Grossart.
Comfit. rend., 1 ( 1 8 4 i ) , 21. 2 3 4 5
+
Dingler's Polytechnisches Journal. 160 (1861), 219. ZeilschvifC f u e r anal. Chemie. 1 (1862), 32. I b i d . , 16 (1877), 267. Zeilschrift fuev p h y s i k . Chemie, 31 (1899), 50.
Sept., 1915
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
a n d cause error in t h e titration, b u t this does not OCcur. T h e presence of too much water in the solution makes t h e reaction lihang,'' i. e., refuse t o begin. If t h e solution is warmed with a few cc. of t h e ferrous sulfate solution, t h e reaction begins abruptly and runs easily thereafter. Heating even t o boiling does not change t h e reaction, b u t m a y volatilize some of t h e HNOI. There should not be more t h a n 2 j per cent of water present a t t h e end of t h e titration. Water added in t h e ferrous sulfate solution must be taken into account. I n t h e light of subsequent knowledge, i t would seem better t o make u p t h e standard solution without a n y sulfuric acid. The proportion of sulfuric acid present in a titration was, however, so small t h a t its influence could not be detected. A few analyses will show t h e accuracy of the method. PER CENT "01 Nitrometer., . . . . . . . . . . . . 0 . 2 5
IN
ARSEKICACID Ferrous sulfate m e t h o d . . .
0.27
Kitric acid was added t o nitre-free arsenic acid and t h e mixture analyzed: Grams nitric acid (1) Taken . . . . . . . . . . . . . . . . . 0.6610 Found.. . . . . . . . . . . . . . . . 0,6608
(2) 0.5968 0,5984
(3) 0.C661 0.0660
A small excess of ferrous sulfate is necessary t o give a good color at t h e end point. Usually, this does not a m o u n t t o more t h a n 0 . I cc., b u t for small quantities of nitrates t h e correction is greater. I n this case a little nitric acid should first be added t o the arsenic acid and titrated with ferrous sulfate t o a brown t i n t ; t h e n t h e sample t o be tested may be added and titrated accurately. ESTIMATION
O F N I T R I C ACID I N P R E S E N C E O F
PHOS-
P H O R I C ACID
T h e same reaction between ferrous salts and nitrates occurs in phosphoric as in arsenic acid, b u t titrations are even more satisfactory. N o heating is required t o s t a r t t h e action and t h e end point is a little sharper t h a n in arsenic acid, being readable t o about 0 . 0 2 cc. with t h e ferrous sulfate solution described above. Practically no excess is required at t h e end point, except for very small amounts of nitrate. As with arsenic acid, t h e concentration of water present a t t h e end of a titration should be less t h a n z j per cent and better results will be obtained with 20 per cent. Below are a few test analyses made in phosphoric acid: GramsHNOa , (1) T a k e n . . . . . . . . . . . 0.6610 Found. . . . . . . . . . . . . . . . 0.6618
(2) 0.5968 0.5984
13)
0,5968 0,5966
E S T I M A T I O h - O F N I T R I C ACID I N P R E S E N C E O F S U L F V R I C ACID
Both arsenic and phosphoric acids have the drawback t h a t t h e y are too expensive for general use in titrations. Sulfuric acid is t h e only strong acid available in t h e laboratory a t a reasonable price. When this was tried t h e surprising fact was discovered t h a t t h e reaction proceeds only two-thirds as far as in arsenic a n d phosphoric acids. Nitrates are reduced, not t o N O , but t o K ? 0 8 . Strong heating will,
767
however, drive t h e reaction somewhat further. reaction t h a t occurs in t h e cold is, therefore: 4FeS04
+ z H N 0 3 + 2H?S04 =
N203
The
+ 2Fe2(S04)3 + 3 H 2 0
The end point is a delicate pinkish brown, not so s h a r p as with t h e other acids, b u t as easy t o read as t h e usual methyl orange end point in acid titrations. E F F E c T 0 F w.4 T E R-E x p eri men t s showed t h a t water slows down t h e reaction and weakens the color of t h e end point, so t h a t a considerable excess of ferrous sulfate is required t o give a readable t i n t . Larger quantities of water stop t h e reaction entirely in t h e cold. Per cent water a t end point 15 20 Gram "01 t a k e n , ,. . . . 0,4963 0.4963 Gram "08 found. . . . . 0.4935 0.4923
20 0.4963 0.5024
25 35 0 , 4 9 6 3 0.5968 0.4973 0.6808
These figures are not of high accuracy, but they show t h a t 2 j per cent of water is permissible as a maxi m u m , though it .makes a rather difficult end point. EFFECT OF TEMPERATURE-The temperature O f titration should be low. Too high a temperature acts in two opposite ways: ( I ) it causes volatilization which leads t o low results; ( 2 ) it causes a of "03, slight reduction of n ' 2 0 3 t o N O , which makes for high values. Maximum temperature SOOC. 6OOC. during titration 40'C. Gram H N O l t a k e n . . . . . 0.4963 0.4963 0.4963 Gram HNOa f o u n d . . . . . . 0.4960 0.4947 0.4940
90'C. llO°C. 0.4963 0.4963 0.5460 0.5038
T h e lowest temperature gives t h e best results; a t medium temperatures t h e compensation of errors is fair b u t irregular; a t high temperatures t h e results are high and the end point is very bad. T h e conclusion is t h a t the temperature should be kept as low a s possible, and should not exceed 60' C. So much heat is generated b'y running t h e ferrous sulfate solution into t h e sulfuric acid t h a t the test beaker must be cooled in a b a t h of water during titration. C O R R E C T I O K F O R E N D POINT-It was found t h a t ferrous sulfate attacks S20a slowly under t h e conditions of the titration, b u t t h a t it has so marked a preference for "03 t h a t no reduction of szO3 occurs while H x o 3 is present. As t h e brown color of t h e end point is due t o t h e formation of a compound of F e S 0 4 a n d K O , some slight reduction t o N O must occur always. T h e results of many experiments show. t h a t about 0 . 2 cc. excess of ferrous sulfate solution ( I cc. = 0 . 0 2 g. "Os) is required t o give the brown t i n t of the end point in a volume of 100-1 j o cc. This holds for titrations ' ranging from 0 . 5 cc. t o a t least 8 0 cc. E F F E C T 0 F C 0 JX V 0 ?j LY OCC LTRR I N G I M P C R I T I E s-- The effect of impurities ordinarily occurring in nitrates was investigated b y adding t h e m t o a known amount of nitric acid and titrating t h e latter. T h e table a t t h e t o p of p. 768 gives t h e results. T h e table shows t h a t nitrite has little effect; chloride, iodide and bromide cause low results in quantities as little a s 2 per cent, and chlorate, bromate and iodate cause high results. Elsewhere in this paper i t is shown t h a t ammonium salts have no effect.
T H E J O U R N A L OF I i V D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
768
INFLUEIVCE OF SALTSOH FERROW SULFATETITRATIONFOR NO, PRESENCE O R SULFURIC ACID Amount of " 0 s taken, 0.3039 g.
IN
SALTSADDEDTO HNOl TITRATION Weight Per cent of Cc. FeSOa Formula Gram mixture tested solution REMARKS None ... Blank.. 15.8 21.9 0.08511 KNOz. 15.71 KKOr . , . . , . 0.4256 :8.6 16.06 ,3.0 K N 0 2 . . , . , . . 0.8511 16.76 Odor of evident 16.0 NaCl. . . . . . . 0.05846 14.96 NaC1. . . . . . . 0.2923 49.0 No distinct end point NaCI. . . . . . . 0.00585 1.9 15:65 11.4 33.2 K I . . . . . . . . . 0.166 5 1 K I . . . . . . . . . 0.017 15.25 0.57 K I . . . . . . . . . 0.0017 15.8 K B r . . . . . . . . 0.1190 28.0 No distinct end point K B r , , , , , , , , 0.0119 li:k6 3.8 KCIOa., , , . . 0.3039 50.0 45.0 N o good end point KIOa. . . . . . . 0.10701 26.0 17.36 K I O J . , . . . . . 0.01070 3.4 15.96 21.6 KBrOa.. . . . . 0.0835 N o distinct end point 2.7 KBrOa.. . . . . 0.00835
..... ......
.
P R 0 C E D U R E R E C 0 M LIE N D E D
Based on t h e foregoing results, t h e following standa r d procedure is recommended for t h e analysis of nitrates with concentrated sulfuric acid as t h e medium: F E R R O G S S U L F A T E soLUTIow-Dissolve 176. j grams F e S 0 4 . 7 H 2 0in 400 cc. water. Stir into this gradually 500 cc. of dilute sulfuric acid, made b y mixing equal volumes of water a n d concentrated acid. Cool t h e mixture a n d make u p t o 1000 cc. The order of mixing given above should be followed, for a different order is a p t t o cause precipitation of a n iron sulfate t h a t cannot be redissolved. 1 cc. = 0.02 g.
"03
and contains 0.8 cc. water
STAh-DARDIZIXG-Either potassium bichromate or nitric acid1 m a y serve for a standard. With t h e former, estimate t h e strength of t h e iron solution b y Penny's method. Remember t o allow 0 . 2 cc. for t h e end point i n titrating nitrates. 1 gram KzCrzO7 = 0.6426 g.
"03
A more satisfactory method of standardizing is t o t i t r a t e a nitric acid solution of known strength under t h e exact conditions in which t h e iron solution is t o be used. For this, dilute 41 cc. of t h e usual 70 per cent laboratory nitric acid t o 1000 cc. a n d titrate with normal causti'c alkali. Use I O cc. of t h e dilute nitric acid t o standardize t h e ferrous sulfate solution in t h e manner described above. A certain solution standardized carefully b y both methods gave t h e following results, showing t h a t t h e methods agree well: CC. = 0.02068g. HNOa (Hh'O, as standard) 1 cc. = 0.02073g. HNOa (KnCrrO7 as standard)
1
TITRATION-The sample should be chosen t o contain 0.3 t o 0 . 6 g. " O s . For t h e most accurate work, t h e titration should be on practically t h e same quant i t y of " 0 3 as was used in standardizing t h e solution. Place I O O cc. concentrated sulfuric acid, free from nitrates, in a z j o cc. beaker set in a large porcelain casserole full of cold water. R u n t h e sample in slowly f r o m a I O cc. pipette t o t h e bottom of t h e acid, stirring meanwhile with t h e pipette; this procedure is designed t o prevent loss of nitric acid fumes. Run in 1 Potassium nitrate may be used for standardizing, b u t i t involves more work than the use of nitric acid, T o d r y a sample of potassium nitrate and test i t for impurities, particularly sodium, is more troublesome t h a n t o titrate a nitric acid solution. The writers met trouble in using a supposedly pure sample of potassium nitrate for standardizing. and for t h a t reason abandoned it in favor of nitric acid.
Vol. 7: No. 9
t h e ferrous sulfate solution slowly in a fine stream with constant stirring until t h e solution turns from yellow t o faint brown or pink. Then rinse out t h e pipette' b y sucking i t full of t h e acid a n d draining it a n d continue titrating cautiously t o t h e first color change. T h e end point is clear t o o . o j cc. and differe n t operators should agree within 0.I cc. T h e casserole of water serves t h e double purpose of cooling t h e acid a n d making t h e end point much clearer. It is sometimes necessary t o halt t h e titration a n d let t h e solution cool. T h e temperature should never exceed 60' C. a n d is better kept below 40' C. Let t h e burette s t a n d five minutes before taking t h e reading, as t h e ferrous sulfate solution drains very slowly. SOME I K D U S T R I A L A P P L I C A T I O N S O F THE METHOD NITRIC
ACID
IN
O L E U M (FUMIKG
SULFURIC
ACID)-
Weigh about I O grams of t h e oleum in a pipette and t i t r a t e according t o t h e scheme given above. A pipette of known content may be used for routine work, so t h a t weighing is unnecessary. This method does not estimate nitrous acid in t h e oleum. T h e accuracy of t h e ferrous sulfate method was given a rigid test b y t w o laboratories b y analyzing t h e same lot of samples independently a n d afterward comparing t h e results. PER CEHT NITRIC ACID IN OLEUM ANALYSES METHOD I I1 I11 IV V Seller's FeSOa.. , . , . 2.40 2.82 3.23 3.35 3.52 Buyer's Nitrometer. 2.35 2.79 3.26 3.39 3.57 DIFFEREKCE. . . . . . . . . . . . . t 0 . 0 5 +0.03 -0.03 -0.04 -0.05 ANALYSES METHOD Seller's FeSOa. . . . . . Buyer's Nitrometer.. D I F F E R E K C E.... . . . . . . . . . .
IX VI1 VI11 VI 3.53 3.57 3.50 3.48 3.50 3.58 3.57 3.53 -0.04 -0.03 -0.02 -0.01
x 3.56 3.56 0.00
The ferrous sulfate method is t h e quickest and probably t h e most accurate way of determining nit r a t e s in oleum. A M ? ~ O N I L T M NITRATE-A test on pure salt showed: Taken.
. . . . . . 0.5000Gram
Found..
. . . . . 0.5004 Gram
The method has been used extensively in one laborat o r y for t h e rapid determination of nitrates in t h e presence of nitrites a n d ammonium salts. This would be a laborious analysis b y t h e usual methods. S O D I U M A X D P O T A S S I U M KITRATES-SampleS O f chemically pure potassium nitrate a n d a sample of commercial nitrate of soda were tested with t h e following results: (2) Commercial iVaSO3 Gram Kh.03 (1) T a k e n , , , . . , , 0.7000 0.3500 By Devarda'smethod F o u n d . . . . . . . 0.6997 0.3510 By FeSO4 method.. . .
97.62per cent 97.65 and 99.13
I n t h e latter case t h e end point was b a d , a n d t h e impurities present caused poor agreement between duplicates. This method is only roughly correct f o r impure nitrate of soda. " M I X E D A C I D " ( N I T R I C A X D S U L F U R I C ACIDS)-samples of mixed acid weighing 6 t o 7 grams were diluted t o exactly I O O cc. with concentrated sulfuric acid and I O cc. portions were titrated. The ferrous sulfate solution used was only one-half a s strong as is recommended above; it was made u p b y diluting one volume 1 The pipette used for measuring the sample should be graduated t o contain, not t o deliver.
Sept., 191j
T H E JOL'RiVAL O F I N D L ' S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
of t h e standard solution with one volume of 60 per cent sulfuric acid. T h e temperature of titration was kept down t o 70' F. b y t h e use of a n ice b a t h . T h e acids used were most carefully analyzed b y t h e evaporation method a n d t h e ferrous sulfate solution was standardized against a mixed acid of known analysis. PERCEXT"01 By FeSOh m e t h o d . . . , . . . . , . . . , . , . . . By evaporation m e t h o d . . , . . , . . . . , .
These analyses tainable b y this serve constancy of I in 300, where
(1) S O . 32 50.15
(2) S O . 16
50.02
(3) 49.49 49.35
represent the limit of accuracy a t method. Reasonable care t o obof conditions permits a n accuracy large amounts of nitrates are present. S G hI M A R Y
I-The authors have devised a method based on t h e well-known ferrous sulfate test for nitrates, b y which large amounts of nitric acid may be titrated directly. 11-The accuracy of t h e method is shown by analyses of known amounts of nitric acid combined and free, comparison being made with the usual standard methods. K i t h care the error does not exceed l/goo of t h e quantity of nitric acid estimated. The method is not suitable for traces. 111-Constancy of conditions is important. The water content should not exceed z j per cent of t h e sample titrated a n d t h e temperature should be kept below 60' C. Chlorates, bromates, iodates, chlorides, bromides and iodides interfere, but nitrites d o not when sulfuric acid is used as a medium. IV-A detailed procedure for analysis is given. \'--Some applications of the method t o technical analysis are included: these are now in regular use in several technical laboratories. In conclusion, t h e authors wish t o acknowledge the cooperation a n d assistance of hilr. W. S. Allen, of the General Chemical Co.. and N r . B. S. Clark. GEKERAI. CHEMICAL COXPANY NEW Y O R E , h-.Y.
A DEVICE FOR THE SUCCESSIVE DETERMINATION OF THE SOLIDS AND FAT IN MILK AND OTHER FLUIDS By ARSIIN SEIDENBERG Received April 27, 1915
A large number of methods for the accurate gravimetric determination of fat in milk have been developed. In one class of methods of which the Rosel-Gottlieb* a n d the T.T'erner-Schmidt3 are examples. t h e proteids are first brought into solution b y the proper reagent, after which the fat is extracted by agitation with a solvent. These methods require a considerable amount of manipulation on t h e part of t h e analyst, t h u s demanding a good deal of his time and attention. On the other hand, t h e direct extraction of the dry residue of the milk in a Soxhlet or modified Soxhlet apparatus, although t h e results are not obtained as quickly, consumes practically no more of the analyst's time t h a n is required t o make t h e necessary weighings.
In order t h a t t h e solvent may act effectively on the d r y milk residue, it is of advantage t o have it in a finely di\-ided state. T o bring this about a number of materials have been suggested, t h u s : sand, pumice stone, asbestos, copper sulfate. plaster of Paris, wood pulp, absorbent paper, etc. Of all these substances the Adamsl fat-free paper offers probably the conditions t h a t are most favorable t o t h e complete extraction of t h e fat since the pores of the paper absorb t h e water in the milk \-ery readily, leaving t h e fat on t h e surface where it is most exposed t o the action of t h e ether. However, there is always some difficulty in transferring a weighed quantity of t h e milk on t o t h e paper. a weighing flask being desirable for most accurate work; also a slight amount of extractable matter remains in t h e paper commonly supplied t o analytical laboratories. When a material is chosen t h a t is not itself hygroscopic. the possibility is presented of making successive determinations of the total solids and fat on t h e same original weighing. In this case the fat can also be determined by difference and t h e solids-not-fat. when these are desired. can be directly weighed instead of being computed from the results of two different determinations. For the determination of total solids and f a t (by difference) and of solids-not-fat, only four weighings are then necessary instead of seTTen, where separate quantities are weighed out for solids and for fat. The Adams fat-free paper is of course n o t adapted t o t h e determination of fat b y difference since it is intensely hygroscopic. The material t h a t has gained most favor for this purpose is asbestos, first proposed by Babcock for the determination of total solids and developed b y RichmondZand also b y hIacFar!ane3 for the determination of fat by difference. As adopted by the Association of Official Agricultural chemist^,^ 1 . j t o 2 . j grams of asbestos are loosely packed into a perforated sheet-metal cylinder 6 0 mm. long and 20 mm. in diameter, closed j mrn. from one end by a perforated disc of t h e same material. T h e total solids are first determined and then t h e fat either b y difference or by weighing directly. T h e asbestos must be prepared first. Freshly ignited woolly asbestos is used from which all grit and small particles must be carefully removed. DESCRIPTIOS
Whatever medium is chosen t o hold the milk, its function is t o permit the ready access ( I ) of heat a t a uniform temperature and ( 2 ) of the solvent used for extraction. 'This is accomplished by using a material t h a t distributes the milk solids in a finely divided state over as large a surface 3s is possible and which, while permitting ready access t o both t h e solvent a n d heat, is acted upon b y neither of these. I t was with these considerations in mind t h a t it occurred t o the writer t h a t a fine meshed wire gauze seemed t o meet all the conditions enumerated, while a t the same time i t offered some advantages in respect t o sim1
1
Zeitschrift angew. Chem., 1888, p. 100; M i l c h . Z t g . , 18 (1888), 264 Landw. V w s . Sla., 40 (18921, 6. Zeifschr. f. analyt. Chemie. 27 (1888). 464.
7 69
2
3
'
Analyst. 10 (1885), 46. Ibid.,17 (1892), 2 2 5 . Ibkd.,18 (1893). 85. L ' . S. Dept. of A g r . , Bureau of Chemistry, Bull. 107, 117