144
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 CHEMISTRY
varies with t h e concentration of t h e solvents employed, t h e proportions of solvent t o material extracted, a n d possibly other physico-chemical .factors, t h e determination of ammonia nitrogen in hydrolyzed flour, flour extracts, a n d gluten should form a basis for a more exact knowledge of t h e proportions in which t h e various proteins occur in flours. Such methods are obviously unadapted t o ordinary analytical use, b u t afford t h e best possible method of exact s t u d y a n d careful investigational work. For example, in Table I1 ( 3 ) , t h e per cent of ammonia nitrogen yielded on hydrolysis of t h e extract with I per cent salt solution indicates clearly t h a t protein other t h a n albumin a n d globulin was extracted, since, as already pointed o u t , t h e per cent of ammonia nitrogen yielded b y pure albumin a n d globulin should be lower. It is suggested also t h a t t o ascertain how much albumin a n d globulin are extracted b y alcohol of a n y given percentage i t should suffice t o hydrolyze t h e alcoholic extract and determine t h e percentage of ammonia nitrogen in a similar manner. Since t h i s article has been in press, t h e results of a s t u d y of t h e purity of proteins extracted from flour by various solvents, using methods here indicated, have been published b y Bailey a n d Blish.’ I n order t o substantiate t h e evidence t h a t there is a relation between t h e ammonia nitrogen yielded on hydrolysis of flour, a n d t h e total quantity of soluble proteins in t h e flours in question, t h e latter were extracted with t a p water (since it was thought desirable t o use t h e same solvent a s was used in washing out t h e glutens) a n d t h e percentage of “soluble nitrogen” in t h e total nitrogen of t h e flour estimated in t h e extract. Although protein material other t h a n globulin a n d albumin is extracted in this process, t h e results should be comparative, a n d in t h e order as indicated above, t h a t is t o say, t h e previously indicated relationship between ammonia nitrogen of t h e hydrolyzed entire flour a n d t h e soluble nitrogen should be apparent. T h a t this is t r u e is indicated b y t h e results obtained, as shown in Table IV. TABLEIV-COMPARISON O F PERCENTAGES O F SOLUBLE NITROGEI I N ENTIRE FLOUR N I T B AMMONIA NITROGEN AFTER HYDROLYSIS OF ENTIREFLOUR Per cent of total N TOTAL Sample No. NITROGEN Soluble N Ammonia N B444 ..................... B440 ..................... B439 ..................... B401 ..................... B452 ..................... B441 B438 ..................... B445 .....................
.....................
2.55 2.17 1.928 2.085 1.917 2.130 1.67 1.26
15.68 18.20 18.67 18.94 19.55 20.42 26.86 28.96
23.00 21.47 21.01 20.81 19.87 21.03 18.85 18.21
Inspection of Table I V shows t h a t , with t h e single exception of B441, as t h e percentage of ammonia nitrogen increases, t h e percentage of soluble nitrogen decreases, a s was expected from t h e theoretical considerations already discussed. CONCLUSIONS
I-The individual proteins of strong a n d weak flours are identical in their chemical constitution, a s determined b y Van Slyke’s method for t h e analysis of proteins. [Since this went t o press, t h e attention of t h e writer 1 “Concerning t h e Identity of t h e Proteins Extracted from Wheat Flour b y t h e Usual Solvents,” J . B i d . Chem., 28 (1915). 345-357.
1’01. 8, S o .
2
was called t o a s t u d y of some of t h e physical constants of gliadin from flours of varying strengths, b y Gr6h a n d Friedl,’ in which t h e y conclude t h a t proteins of different flours have t h e same constitution.] 11-The ratio of gliadin t o glutenin is much more nearly constant in flours of different baking qualities t h a n has heretofore been supposed. 111-There is a far greater variation in t h e percentages of t h e so-called “soluble proteins” (albumin a n d globulin) in flours. IV-Since t h e various proteins in t h e same flour differ widely in their content of ammonia nitrogen, t h e determination of ammonia nitrogen in flours, in extracts of flours made with various solvents, a n d i n t h e crude gluten of flours, after their previous complete hydrolysis with strong mineral acid, can be made t o serve a s a n accurate indication of t h e amounts of t h e various proteins present, since t h e proteins of widely different flours have been shown t o have t h e same chemical constitution. Acknowledgment a n d sincere t h a n k s are herewith extended t o Professor R. W. Thatcher under whose supervision this work was done, also t o Dr. R. A. Gortner a n d Professor C. H. Bailey for many valuable a n d helpful suggestions. DEPARTMENT O F AGRICULTURE UNIVERSITYOF MINIESOTA ST. PAUL, MINXESOTA
THE ANALYSIS OF MAPLE PRODUCTS, V Miscellaneous Observations on Maple Syrup Incidental to a Search for New Methods of Detecting Adulteration2 B y J. F. SNELL Received August 20, 1915
I n t h e course of a search for new a n d improved methods of detecting adulteration in maple syrup, t h e author a n d his associates have made a number of miscellaneous observations which may be suggestive or otherwise useful t o other investigators in this line. I-EFFECTS
OF SOME REAGENTS
It is well known t h a t lead subacetate produces a heavy precipitate in maple syrup a n d t h e normal acetate one not so heavy. I n t h e “malic acid value” determination, alcohol a n d calcium chloride are employed together as precipitants. I t appeared t o t h e author not improbable t h a t a systematic search would reveal other precipitating reagents whose behavior might prove equally useful in t h e examination of maple products. Mr. J. M. Scott made some desultory experiments in this direction and has recorded in his notes “indifferent” results with ether, zinc acetate, uranium acetate, cadmium chloride, bismuth nitrate, copper nitrate and cupric chloride. Mercuric acetate gave a flocculent precipitate which settled on standing and was light yellow in color. Silver nitrate in 50 per cent solution gave a white precipitate which darkened on standing. T h e author has also observed t h a t barium salts give no precipitate in maple syrup although t h e y do so in solutions of t h e Biochem. Zlschr., 66 (1914), 154. Presented a t t h e 51st Meeting of t h e Amencan Chemical Society. Seattle, Aug. 31-Sept. 3, 1915. Previous papers of this series-Tars JOURNAL,6 (1913), 740, 993; 6 (1914), 216, 301. 1
2
Feb., 1916
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
ash. Alcohol b y itself gives a precipitate containing calcium, potassium a n d organic matter. By far t h e greater p a r t of t h e total calcium a n d of t h e total potassium of t h e syrup are precipitated b y t h e addition of a suitable q u a n t i t y of alcohol. It was Mr. Scott’s intention t o investigate this alcohol precipitate further, b u t his removal t o another position prevented his returning t o t h e point. 11-EXPERINENTS
WITH SILVER NITRATE
Silver nitrate having been found t o give a decided precipitate in maple syrup a n d none in a syrup made from white sugar, it was thought t h a t a “silver number” might be determined b y a method analogous t o t h e Winton method with lead subacetate. Mr. Scott made a number of experiments towards this end. T h e idea was t o a d d a n excess of a s t a n d a r d solution of silver nitrate, filter off t h e precipitate, determine t h e residual silver volumetrically, a n d calculate from t h e results a “silver number” representing t h e quantity of silver precipitated b y I O O grams of t h e syrup (or, alternatively, b y I O O grams of t h e dry matter of t h e s y r u p ) . It was found, however, t h a t t h e longer t h e solution containing t h e excess of silver was allowed t o s t a n d (and standing appeared t o be necessary t o allow t h e precipitate t o collect) t h e lower was t h e q u a n t i t y of residual silver. I n other words, precipitation of silver continued during a period of several hours. For example, in one series of experiments 50 cc. N / r o silver nitrate were added t o 25 cc. syrup a n d t h e mixture made u p t o roo cc. Filtered 2 0 cc. aliquots, t a k e n , respectively, after I , 2, 4 a n d 6 hours, required t h e following quantities of N/IO sodium chloride t o precipitate t h e residual silver: 8.00 cc.. 7.55 cc., 7.40 cc. a n d 7.20 cc. I t is conceivable t h a t a method using silver nitrate might be worked o u t , b u t obviously this circumstance of a slow reaction in t h e solution resulting in t h e precipitation of silver would have t o be t a k e n into account. T h e slow reaction would appear t o be worthy of s t u d y also for its theoretical interest. Has i t a n y connection with t h e darkening of t h e precipitate? I s i t a reduction? W h a t constituent of maple syrup is it which reacts t h u s with silver nitrate? Some experiments with a N / j o silver nitrate solution, made b y Mr. Tu’. C. MacFarlane, are reported in Sections V a n d V I below. 111-A
LEAD
SUBACETATE
METHOD
BASED
ON
THE
W I X T O N AKD CANADIAN M E T H O D S
I n Paper I1 of this series’ reference was made t o some experiments in which a n endeavor was made t o combine t h e advantages of t h e Canadian a n d t h e Winton methods of detecting adulteration of maple s y r u p b y t h e use of lead subacetate.2 T h e outcome of these experiments, made b y Mr. Scott, was a method i n which 2 5 grams syrup were treated with I O cc.3 of lead subacetate solution of sp. gr. 1 . 2 5 , enough water added t o bring t h e volume t o roo cc. a n d , after Snell and Scott, THISJOURNAL 6 (1913), 993. Loc. rzt., p. 996. * Experiments were also made with 15 cc. of t h e lead subacetate solution. The range of variation of the lead values so obtained was, however, 464 per cent of the minimum as compared with 200 per cent in t h e method with 10 cc. 1 2
I45
standing three hours a n d filtering, a I O cc. aliquot taken f o r determination of unprecipitated lead as in t h e Winton method. Comparison with a blank, using a syrup made from granulated sugar,l enabled one t o calculate t h e amount of lead precipitated b y I O O grams of t h e maple syrup. This q u a n t i t y of lead was t o be t h e “lead number” of t h e proposed method. For. comparison with t h e Canadian a n d Winton methods this proposed lead number was determined upon t w e n t y of t h e syrups whose analysis was reported in Paper II12--viz., t e n of those which gave high, a n d t e n of those which gave low, Canadian a n d “modified” Winton lead value^.^ T h e values obtained b y t h e projected method were in all instances lower t h a n those given b y t h e modified Winton method. T h e ranges of variation of t h e three lead numbers in these t w e n t y syrups, expressed in percentage of t h e minimum, were found t o be: Per cent Canadian Lead Numbe . . . . . . . . . . . . . . . . . . 190 Modified Winton Lead Proposed Lead Number.. . . . . . . . . . . . . . . . . . . . . . 200 ( a ) T h e syrup which gave t h e highest Canadian lead number was not included among t h e twenty. Hence t h e range of this value among the twenty syrups is a little smaller than t h a t among the 126 reported in Paper 111.
.
.
Inasmuch as the lead number determined b y t h e original Winton method (using 25 grams syrup a n d not reducing t o t h e dry basis) has a narrower general range of variation t h a n the value obtained b y reducing t h e results t o t h e dry matter basis,4 it appeared not improbable t h a t if t h e comparison were made between t h e true Winton a n d t h e proposed method, t h e advantage in favor of t h e former would be more considerable t h a n t h a t in favor of t h e modified Winton number. Determination of t h e t r u e Winton number upon nine of t h e twenty syrups corroborated t h i s conjecture, t h e ranges of t h e values among these nine syrups being as follows: Per cent of minimum Canadian Lead Number . , , , , ,,.,,. . . . 192 149 Modified Winton 149 Proposed Lead N 119(a) True Winton Lea ( a ) This is exactly the same range of variation as is shown in McGill’s cit., p. 221). Bryan’s analysis of 47 samples (see Paper 111, Table VI, LOG. results on 481 samples showed a range only slightly higher, viz., 135 per cent (Loc. &.). B u t if McGill’s results and Bryan’s are taken in conjunction, the true Winton value shows a range of 286 per cent of t h e minimum value. One cannot help suspecting, however, t h a t there was some difference in manipulation in t h e two laboratories which resulted in the Canadian obtaining uniformly lower results than the American.
.
......
. . .. ..
These results indicate t h a t t h e proposed lead method, although having a decided advantage over t h e Canadian lead method in t h e matter of t h e range of t h e value i n pure syrups would not be t h e equal of t h e Winton method in this r e ~ p e c t . ~ IV-FALLING
OFF
OB VALUE
D U E TO DILUTIOPT
WITH
SUCROSE
SYRUP-If t h e proposed new lead subacetate method proved t o yield a value which shared with t h e Canadian C i t . , p. 997. Snell and Scott, THIS JOURNAL, 6 (19141, 216. 8 Using t h e quantities of syrup containing 5 grams and 25 grams dry matter, respectively, as prescribed in t h e Canadian pure food standards. 4 See Table V I of Paper 111, LOG. cit., p. 221. 5 The author desires t o reiterate his opinion t h a t t h e original Winton metbod, without reduction t o t h e dry basis, is better than any of the modifications of i t which have been proposed. Comparison with a blank is, however, essential, and the suggestion is here repeated t h a t a syrup made from (commercial) pure cane sugar be employed in the blank determination. See Paper 11, LOC.cit., p. 997. 1 LOG. 2
T H E J O U R L V A L O F I K D C S T R I A L A N D EAiGINEERIiV'G C H E M I S T R Y
136
lead value the advantage of falling off rapidly when the maple syrup is adulterated with a sucrose syrup of the same density; it might be worth recommending in spite of its apparent inferiority t o the true V i n t o n lead number in the matter of range of variation. The diminution of the new lead value was compared with t h a t of the Canadian and true TTinton values in three syrups, determinations of all three lead numbers being made on the pure syrups and on mixtures of each with sucrose syrup in such proportions as t o give syrups containing 90, 80. 70: jo and 30 per cent of maple syrup. I n t h e first of t h e three syrups there was little choice among the three methods except in the mixture containing only 3 0 per cent of maple. For this mixture the Canadian lead value mas only 18.7 per cent, the K i n t o n value was 2 2 . 1 , and the proposed lead number 33.3 per cent of t h e corresponding values for t h e pure maple syrup. I n other words, t h e proposed lead method showed itself inferior t o either of t h e old methods. I n the other t w o syrups t h e proposed method showed a distinct advantage over the T i n t o n method and for some of the mixtures appeared t o be even slightly better t h a n the Canadian method. On the whole this projected lead value had compared sufficiently favorably with t h e established methods t o justify further investigation, had we not hit upon another lead subacetate method which appears much superior t o a n y hitherto proposed and which will form the subject of the next paper of this series. V-LEAD
COKTENT
OF
THE
"CANADIAN
LEAD'!
PRE-
CIPITATE
hlr. J. 11. Scott determined the quantity of lead in the Canadian lead precipitates from 8 samples of maple syrup, representing a wide range of Canadian lead number. The precipitate was washed as directed in the Canadian lead method, i. e . , four or five times with hot water. I t was then dried, weighed and subsequently decomposed with aqua regia, after which the lead was converted into lead sulfate in the usual way and weighed as such. The results obtained are shown in Table 1. TABLE I-"
CANADIAN
LEAD"P R E C I P I T A T E S
SOGRCE
NU I
OF SYRUP
DESCRIPTION Canada Composite of 54 syrups Canada Mixture of several syrups Middle run, pan, white of egg Dutton, Ont. Brome, Que. Early run, pan Hillsburgh, Ont. Early run, evaporator Wardsville, O n r . Early run, evaporator Orangeville, Ont. Middle run, evaporator, milk Neustadt, Ont. Ear!? run. kettle, eggs A4V3RhGE OP ~ ~ I J M B E R 3 ST O 8,
Canadian Per cent lead P b in h-0.
3.44 4.40 1.74 2.86 2.01 2.33 5.08 4.10 3.02
D D t. . _
69.41 70.11 69.62 68.65 68.51 68.50 68.25 66.95 68.415
The results indicate a fairly definite composition in t h e precipitate. the variation in lead content being just over 3 per cent of the total weight. There appears t o be no definite relation between the quantities of precipitate yielded by the syrup (Canadian lead numbers) and the composition of t h e precipitates. I n vie117 of the fact t h a t this precipitate has sometimes been erroneously assumed t o consist of normal lead malate, it is m-orth noting t h a t the latter salt in t h e anhydrous conditioz contains only 6 1 . 0 0 per cent of
ITol.8. KO.2
lead. I t is possible t h a t a basic malate' may be t h e essential, or a t all events an important, constituent of the precipitate. but evidence t o t h a t effect has not yet been adduced. I t is our intention t o make a s t u d y of the organic non-sugar constituents of maple syrup which will embrace a more complete analysis of t h e lead precipitate as well as of the other precipitates obtainable from maple syrup and not from syrups made from pure sucrose. VI-ATTEJIPTS
TO DETERhIIKE IKORGANIC COXSTITUESTS
V 0 LU 31E T R I C A L L Y
If among the non-sugar constituents of maple syrup one could be found showing a reasonable degree of constancy in all genuine syrups, a coni-enient method of determining that constituent would obviously constitute a useful method of detecting adulteration.' I t appeared a possibility t h a t among the inorganic constituents one might be found answering t h e requirements. The recorded analyses of the ash of maple syrup and sugar are comparatively few and the only complete analyses known t o the writer are those of Hortvet,3 embracing only two sugars and two syrups. The constituents which have been determined in a larger number of samples-uiz., potash, lime, phosphoric acid and sulfuric acid-all show very wide variations in individual syrups. Jones4 found the potash in a maple sugar ash as l o w as 18.26 per cent while Bryan' obtained results in syrups ranging from z 3 . j j t o j q . j q . For lime Bryan obtained results running from 13.20 t o 36.36; for phosphoric acid, 1.08t o 12.90; for sulfuric acid, 0 . 0 0 t o 6.18 per cent. The results obtained by the other in\-estigators all fall within t h e above limits. Mr. J. AI. Scott in 1913 made complete analyses of the soluble and insoluble portions of the ash of a syrup compounded of about sixty genuine Canadian syrups. This analysis will probably be published in a future paper of this series dealing with the ash. From its d a t a and the ratio of t h e soluble t o the insoluble ash in t h e same composite syrup (which ratio happens t o be unity) t h e composition of t h e total ash can be calculated. I t is as follows: EL0 . . . 29.0; MgO . . . , 2 . 3 4 COS. . . . , . 3 3 , 7 5 SOa.. , . . . . . , . 1.37 hTasO.. 1 . 9 7 Fez03.. . . 1 . 8 2 SiOz.. . . , 1 . 2 2 C I . , , , , . . . . . . 1 . 4 6 ,
C a O . . . 2 5 . 8 8 RZnaO?:. , 0 . 8 1 P s O ~ ,, ,, . . 0 . 1 5 Insol. in Hl?.. 0.1: TOTAL. after deducting oxygen-equivalent of t h e chlorine, , , . . . . . 9 9 . 5 8
I t will be noted t h a t the quantity of phosphoric acid found is only one-seventh of the minimum percentage hitherto reported, t h u s extending the range of variation for this constituent. I t is clear, then, t h a t none of the four ash constituents which have been extensively studied is suitable 1 Otto's basic lead malate (Liebig's A n n . , 117, 177) contained i 3 . 8 per cent of lead, while t h e salt he obtained b y adding t h e neutralized or nearly neutralized acid t o a boiling excess of lead acetate solution contained 7 5 . 8 3 per cent of lead in its anhydrous condition, which i,5 equivalent t o 7 3 . 5 per cent in t h e monohydrate, which Otto found t o be stable a t 100° C. 2 According t o Hortvet, t h e protein (N X 6 . 2 5 ) content is fairly cons t a n t and has sume value in this direction. So far as t h e writer is aware, b u t little attention has been paid t o this suggestion of Hortvet's. 3 Hortvet, J . Am. Chem. SOL.,26 (1904). 1541. 4 Jones, Vermont Bgr. Expt. Station, 1811~Annual Repoul, 1904-6, p . 331. 5 Bryan, Bureau of Chem., U. S. Dept. Agr., BzrZZ. 194, pp. 82-89.
Feb., 1916
T H E J O U R N A L O F I N D U S T R I A L A N D En'GINEERING C H E M I S T R Y
for our purpose. Magnesia shows less variability (2.34 t o 4.6 j ) in t h e five instances in which it has been determined, b u t no easy method of determining it in presence of the other constituents of t h e syrup suggests itself. . The fact t h a t Hortvet found extremely small quantities of chlorine in t h e ashes he analyzed suggested t h a t a chlorine determination might be useful, particularly in case raw cane sugar or refined molasses were employed in adu1terating.l h4r. Scott's results, however, showing a very material chlorine content in a composite of so many samples, rather discouraged our hopes a s t o chlorine. Nevertheless, some experiments in titrating t h e diluted syrups with N / j o silver nitrate were made b y M r . S . C. MacFarlane. Quantities of syrup containing I O grams of solid matter were treated with zoo cc. water, IO cc. lead subacetate solution (sp. gr. 1 . 2 j ) and I O cc. saturated potassium chromate, t h e assumption being t h a t the lead would precipitate the organic matter precipitable b y silver and at t h e same time decolorize t h e liquid, t h u s enabling t h e potassium chromate left in solution t o act as a n indicator in t h e titration for chlorine. The precipitate was washed with hot water and t h e filtrate (including washings) titrated with N / j O silver nitrate. Xine genuine syrups took from 0.82 t o 4.60 cc. of the silver nitrate. T h e same method, applied t o 3 samples of raw cane sugar, gave 4.32, 5.14 and 9.08 as the number of cc. of silver nitrate required €or I O grams d r y matter. A sample of beet sugar crystals gave 1.94and a sample of cane molasses 51.99. The method, accordingly, offers little advantage except in t h e case of adulteration with molasses. Experiments were also made in which t h e syrups were treated with alumina cream instead of lead subacetate. Syrup containing I O g. solid matter was treated with zoo cc. water and I O cc. alumina cream. T h e precipitate was washed with cold water and t h e filtrate and washings titrated with iV/jo silver nitrate, using potassium chromate as indicator. The same nine pure syrups as before took 1.01 t o 6.40 cc. of t h e silver nitrate solution. I n 1 2 other genuine maple syrups results were obtained ranging from 2.80 t o 8.38. T h e raw cane sugars took 9.85, 3.64 and 7 . 5 2 cc. respectively, t h e beet sugar 2.23 and t h e molasses j 7 . 0 0 . The results b y t h e two methods did not r u n parallel. I n general, more silver nitrate was required after the alumina cream t h a n after t h e lead subacetate treatment, b u t there were exceptions t o this rule. I n view of t h e probability t h a t the phosphorus in maple syrup is largely in organic combination, it was thought t h a t in spite of the wide variability of t h e phosphoric acid content of t h e ash, the amount of phosphorus in t h e syrup itself precipitable by uranyl acetate might prove to be reasonably constant. Mr. MacFarlane made some titrations which appeared t o indicate a fair degree of constancy in this value in t h e same nine syrups as were used in his experiments with silver nitrate. T h e results obtained with t h e 1 Practical men in Canada appear t o be of the opinion t h a t raw sugars or molasses are n o t commonly used a- adulterants, b u t I have not thought i t advisable t o ignore t h e possibility.
=47
raw sugars, however, were very similar t o those with t h e maple syrups, I t is doubtful whether the method employed-titration of 20 cc. syrup with a uranyl acetate solution of which I cc. is equivalent t o 5 mg. of P206-was suitable for the measurement of such small quantities of phosphoric acid and i t is possible t h a t further investigation along this line might be useful. VII-ELECTROTITRAMETRIC
EXPERIMENTS
T h e interesting paper of'F. H. Hesselink van Suchtelen a n d Arao Itano,' pointin'g out t h a t measurements of electrical resistame might be employed for t h e detection of t h e end-point in precipitation titrations, suggested t o us t h a t we might titrate syrups with silver nitrate without previous treatment with lead subacetate or alumina cream, using the electrical resistance as indicator in place of potassium chromate. I t was, indeed, found possible t o get a definite endpoint in this way-indicated b y a break in t h e curve on t h e plot of electrical resistance with volume of silver nitrate added: 3 cc. syrup were diluted t o 30 cc. and IV/ j o silver nitrate added from a burette, the electrical resistance being measured with a dip electrode after the addition of each cc. of the reagent. Fifteen genuine syrups gave end-points a t 0 . 2 0 t o 1.9j cc. silver nitrate. The range of variation is, therefore, too wide t o render t h e method useful. Similar experiments were made with N / I O barium nitrate solution. With this reagent, as might be anticipated from t h e fact t h a t no precipitate is produced, the resistance-volume curves showed no break. I t then occurred t o us t h a t the electrical resistance might be employed t o detect t h e end-point in a titration with lead subacetate solution. Mr. MacFarlane, who had conducted t h e experiments reported above, made a number of experiments in this line, finally adopting a procedure in which a solution of lead subacetate prepared by diluting a solution of sp. gr. 1.246 ( a t I j " C.) with ten times its volume of water2 was added from a burette t o 3 cc. syrup diluted t o 30 cc. This method applied t o 14 samples of pure syrup gave results varying from 4.5 t o 6.8 cc. of t h e lead subacetate solution-a range of j I per cent. Later Mr. G. J. Van Zoeren applied this method t o 24 syrups of the season of 191j , obtaining results ranging from 4.0 t o 5.4. The total range among t h e 38 syrups is, accordingly, 7 0 per cent of the minimum. The dip electrode used in the above experiments was of one of the common forms made €or use with poor conductor^.^ A new electrode was designed by M r . T'an Zoeren which can be used for both t h e conductivity test described in Paper 1 of this series and the volumetric lead subacetate test t o be described in Paper VI. A description of this electrode will be published in the Jourizal of the A m e y i c a n C h e m i c a l Society. T h e new electrode being larger t h a n t h a t previously used it became necessary t o employ a larger volume of liquid t h a n Mr. MacFarlane had used: 60 cc. were 2
3
v a n Suchtelen a n d Itano, J . A m . Chem. Soc., 36 (1914), 1793. This solution has a density of 1,025. Baird a n d Tatlock's Catalogue, 1914, No. 5313T.
148
T H E J O U R N A L O F I N D U S T R I A L A N D ENGFNEERING C H E M I S T R Y
found t o be a suitable quantity of the diluted syrup for the purpose. Experiments were then made in which t h e concentration of t h e syrup and subacetate solutions were varied. T h e results are shown in Table 11. T h e procedure adopted on t h e basis of these experiments will be described in the next paper. Applied to t h e same 24 syrups of t h e season of 1915 as were tested by M r . MacFarlane’s procedure, the Van Zoeren procedure gave results ranging from 4.8 t o 6.0. This range amounts t o 2 2 per cent of the mean (5.55) or 2 5 per cent of the minimum as against 30 per cent of the mean (4.70) or 35 per cent of the minimum for t h e LIacFarlane procedure. TABLE11-TITRAMETRIC EXFERIMEKTS WITH LEADSUBACETATE Range of lead numbers Cc. syrup No. of I n per Density of subacetate diluted syrups Actual cent of solution t o 100 cc. tested minimum 1.050 10 6 KO definite breaks 15 23 2.5-4.5 80 20 8 2.4-4.7 96 1.033 5 16 I n 5 instances no breaks 8 12 4.0-5.9 41 10 16 4.7-6.5 38
SUMMARY
I-Silver nitrate added t o maple syrup gives a white precipitate which darkens on standing. The precipitation of silver continues during a period of several hours. 2-Mercuric acetate added t o maple syrup produces a light yellow precipitate. 3-Alcohol produces a precipitate containing most of t h e calcium and potassium. 4-4 moderately successful a t t e m p t was made t o combine t h e advantages of t h e Winton and Canadian lead subacetate methods. 5-The Canadian lead precipitates from six syrups showed a lead content of 66.95 t o 69.62 per cent; average, 68.42. The precipitate from a composite of 54 syrups contained 69.41 per cent of lead, while t h a t from another mixed syrup contained 70.1 I per cent. 6-Titration of maple syrup with N,’jo silver nitrate ( I ) directly, using electrical resistance measurements t o detect the end-point, ( 2 ) after treatment with lead subacetate or alumina cream, using potassium chromate as indicator, yielded definite b a t not useful results. 7-Titration with uranyl acetate gave no useful results. 8-Titration with lead subacetate solutions using electrical resistance as indicator led t o a useful method of testing the syrup for purity, which is described in t h e next paper of t h e series. 9-A complete anaIysis of t h e ash of a composite of about sixty genuine syrups shows more chlorine and less phosphoric acid t h a n t h e analyses previously recorded. MACDONALD COLLEGE, QUEBEC, CANADA
A COMPARISON OF METHODS FOR THE DETERMINATION OF SOIL PHOSPHORUS By W. 0. ROBINSON Received August 7, 1915
Hillebrandl outlines the main points of this determination in t h e following statement: “ I t is some1
U. S. Geol. Survey, B d l . 422 (1910), 144.
Vol. 8, NO. 2
times possible to extract all the phosphorus from a rock b y simple digestion with nitric acid, b u t quite as often if not oftener this fails; hence t h e necessity of resorting t o one of t h e longer methods of extraction 9 x x . Whatever method is used, great care is required t o secure accurate results. ” This statement applies t o soils with t h e further complication t h a t organic matter is inlTariably present, sometimes in amounts large enough t o require special treatment. METHODS O F S O L U T I O N O F T H E S O I L
Washington1 and P r y Z have called attention t o the presence of apatite inclusions in quartz as affecting the determination of phosphoric acid. Protected in this way the apatite would not, of course, be soluble in acids other t h a n hydrofluoric. Further, there are many phosphate minerals such as variscite, wavellite, and xenotime, which, from the standpoint of determinative mineralogy, are classed as insoluble in acids. These minerals have not been reported as occurring in soils, b u t there is a possibility of it. F r y 3 cites a number of analyses showing t h a t in most cases the total phosphoric acid is not dissolved b y acid digestion: from 4 t o I O O per cent is extracted. Obviously a method of simple acid digestion will not be generally reliable when applied t o soils. F I S C H E R ’ S i t r ~ ~ ~ o ~ - F i s c h ehowever, r,~ has modified a n acid digestion method so t h a t it appears t o extract t h e entire amount of phosphorus. The salient feature of this process is a n intervening ignition between two acid treatments. Briefly t h e procedure is as follows: j-10 grams of soil are treated with 5 0 cc. of aqua regia in a covered quartz dish of appropriate capacity. After t h e action has ceased t h e cover is removed and the mass evaporated t o dryness. It is then ignited (ostensibly long enough t o destroy organic matter) and again treated with aqua regia, evaporated t o dryness and taken up with nitric acid. Fischer claims t h a t the process gives slightly higher results t h a n the fusion method,’ though he proves there is a n almost negligible amount of phosphorus left in the insoluble residue. He points out t h a t a larger sample can be conveniently used with his method t h a n with t h e fusion method, thereby securing a fairer sample of t h e soil. The author has tested this method on a variety of soils. T h e samples were well ground a n a mixed so t h a t b u t one gram was employed in each case. After the second aqua regia treatment, t h e mass was evaporated once with nitric acid, heated on the hot plate t o browning t o dehydrate t h e silica, then taken u p with the requisite amount of nitric acid. T h e results, and for comparison, those obtained by other methods, are given in Table I. Fair agreement is shown, considering t h e determina1 “The Chemical Analysis of Rocks,” Wiley & Sons, Xew York (1910), p . 162. 2 THISJOURNAL, 6 (1913), 664. 8 L O C . cit. 4 Intern. Mitt. J’. Bodenkunde, 2 (1913), 541. 6 Probably due to the difficulty of precipitating small amounts of ammonium phosphomolybdate in presence of much NaNOa in the fusion method since Cajn and Hostetter [ J . Soc. Chem. l a d . , 4 (1912), 2501 have shown t h a t aqueous solutions of NaNOs have a strong solvent effect on the phospbomolybdate when vanadium is present.
