Mar., 1917
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
and flooding an excess of phenoldisulfonic acid quickly over the salt. j-All nitrates in a soil solution can be recovered regardless of the salts present therein. &Potash alum may be used as a flocculent in preparing t h e soil solution without producing a loss of nitrates. By the old method the loss of nitrates in the presence of certain salts was often as high as 50 per cent. 7-Since potash alum is an excellent flocculent, soil chemists and soil bacteriologists need not hesitate t o employ its use, provided they use the modified phenoldisulfonic acid method. BIBLIOGRAPHY Schreiner anc Failyer, “Colorimetric, Turbidity and Titration Methods Used in Soil Investigations,” Bureau of Soils, U. S. Dept. Agri., Bull. 1
31. f C. B. Lipman and L. T. Sharp, “Studies on the Phenoldisulphonic Acid Method for Determining Nitrates in Soils,” Unio. Cal. Pub. i n Agr. Sci., Vol. 1, No. 2. :E. M. Chamot and D. S. Pratt, “A Study of the Phenoldisulfonic Acid Method for the Determination of Nitrates in Water-The Composition of the Yellow Compound,” Jour. A m . Chem. Soc.. Vol. 32, pp. 630-637. 4 E. M. Chamot. D. S. Pratt and H. W. Redfield, “A Study on the Phenoldisulfonic Acid Method for the Determination of Nitrates in Waterthe Chief Sources of Error in the Method,” Jour. A m . Chem. SOC.,Vol. 83, pp. 366-381. I E. M. Chamot, D. S. P r a t t and H. W. Redfield. “A Study of the Phenoldisulfonic Acid Method for the Determination of Nitrates in Water-A Modified Phenoldisulfonic Acid Method,” Ibid., Vol. 33, pp. 38 1-384. 8 H. H. Hill, “The Determination of Nitrates in Soil and Soil Extracts,” Annual Reporl of lhe V a . E x g l . Sla., 1911-12, p. 133. W P. Kelly, “The Effects of Sulfates on the Determination of Nitrates,” Jour. A m . Chem. SOC.,36 (1913), 775. Robert Stewart, “The Occurrence of Potassium Nitrate in Western America.” Jour. A m . Chcm. SOC.,33 (1911). 1952. R. Stewart and J. E. Greaves, “The Influence of Chlorine on the Determination of Nitrates by the Phenoldisulfonic Acid Method,“ Jour. A m . Chem. Soc., 36 (1913). 579. 10 R. Stewart and J. E. Greaves, “The Influence of Chlorine upon the Determination of Nitric Nitrogen,” I b i d . , 34 (1910). 756. 11 E. M. Chamot and D. S. Pratt, “A Study of the Phenoldisulfonic ComposiAcid Method for the Determination of Nitrates i n Water-The tion of the Reagent and the Reaction Product,” Jour. A m . Chcm. SOC., 31 (1909). 922. 12 A. H. Gill, “ O n the Determination of Nitrates in Potable Water,“ Jour. A m . Chem. Soc., 16 (1894). 122. 18 King and Whitson. WIS.Agri. Exp Sta.. Bulls. 86 and 93. 1 4 Baer, Thesis (unpublished) for M.S.in Agriculture, Univ. of Wisconsin (1914). Tiemann-Gaertner “Wasseranalyse,“ 3rd Ed., p. 168. 11 Schulze-Tiemann, Ber. d . chem. Gcs , 6 , 1041. 17 Schl6ssing-Reichardt. Z . anal. Chem.. 9, p. 24. 1: Crum-Lunge. Phil. Mag.. [3] 30, 426. 1) Marx-Trommsdorf, 2. anal. Chem , 9, 171. 10 Sprengel, Pogg. Ann., 121, 188. (1 Grandval and Lajoux, Comfit. rend.. 10, 1, 62. I* Fox, Tech. Quart.. 1, 1. 1) Johnson,Chem. News, 61, 15. “Lind, Chem. News, 68, 1, 15, 28. Smith, Analyst. 10, 197. Bartram, Jour. Frank. I n s f . . March‘l7, 1891. Hazen and Clark, J . Anal. Appl. Chem., 6, 1. 1: E x p . Sto. Record, 6, 404. 2) Fox, Tech. Quart., 1, 54. 80 Andrews. Jour. A m . Chcm. Soc.. 26, 388. Montenari, Gazz. chim. ital., 32, 1, 87, 1902. I* Leeds, A m . Jour. Sci.. [3] 7 , 197. ‘1 Weston, t a b . Uhiv. of Ill, 1909. Mercille, A m . Chem. Anal., 14 (1909). 303. Ulsch, New Jersey E W . Sta. Rcpl., 1892,~188-193.
*
*
WEST
T B N N I W SST~A~T I ~ ~ N O R Y ASCHOOL L MEMPHIS.TENNBSSRB
295
A NOTE ON THE DETECTION AND ESTIMATION OF SMALL AMOUNTS OF METHYL. ALCOHOL By ELIASELVOVE Received January 24. 1917
I n applying the DenigBsl test for methyl alcohol
t o its colorimetric estimation, Simmondsz recommends t h a t the solution be always made up t o contain I O per cent3 ethyl alcohol. I n studying this method with the view of applying it t o the detection and estimation of methyl alcohol vapor in air it was found, however, t h a t the test can be made more sensitive b y reducing the proportion of ethyl alcohol from I O t o 0.5 per cent. T h a t this is so may be seen from the following experiment:‘ T o 2 . 5 cc. of an aqueous solution of methyl alcohol containing 0.3 mg. methyl alcohol (No. I), in a 5 0 cc. Erlenmeyer flask, there were added 2.5 cc. of a one per cent aqueous solution of ethyl alcohol. T o another equal volume of this methyl alcohol solution (No. z ) , there were added 2.5 cc. of a 2 0 per cent aqueous solution of ethyl alcohol. Each solution was then treated with 2.5 cc. of a 2 per cent potassium permanganate solution and 0.2 cc. concentrated sulfuric acid and allowed to stand 3 minutes The excess of the permanganate was then reduced by mixing each solution with 0.7 cc. of6 a 9.6 per cent6 oxalic acid solution. Each solution was then further acidified by mixing with one cc. 6f concentrated sulfuric acid, allowed to cool t o room temperature, and finally mixed with 5 cc. of Schiff’s reagent. After standing 40 minutes, they were compared in the narrow form jo cc. Nessler tubes and i t was found t h a t whereas No. I had developed a very decided color, No. 2 was almost completely colorless in comparison. After standing an hour, this difference was even more pronounced, the color 1
Comfit. rend., 160 (1910), 832.
* Analysf. 37 (1912).
16. All percentages of ethyl alcohol mentioned in this paper refer to percentage by volume. 4 The details of this experiment also give the essentials of the procedure which was finally adopted. To apply this for estimating the methyl alcohol In its aqueous solution, such as may be obtained by suitably passing through water a definite volume of air carrying methyl alcohol vapors, proceed as follows: Ascertain by a preliminary experiment the approximate amount of methyl alcohol in the solution. If this shows t h a t 5 cc. of it contain more than 1 mg. of methyl alcohol, dilute so as t o bring it within this limit. Mix 4.5 CC. of this diluted solution with 0.5 cc. of 5 per cent ethyl alcohol. Similarly prepare several 5 cc. portions of methyl alcohol solutions by diluting the proper amounts of a 0.1 per cent (0.1 g. to 100 cc.) aqueous solution of methyl alcohol to 4.5 CC. with water and then adding 0.5 CC. of 5 per cent ethyl alcohol t o each. These standard methyl alcohol solutions are made t o vary by 0.1 mg. methyl alcohol and the limits are chosen so as t o bring the unknown within their range. The unknown and the standards are then subjected exactly alike, preferably in 50 cc. Erlenmeyer flasks, to the permanganate treatment and the subsequent mixing with Schiff’s reagent as described above. The solutions are finally transferred into the narrow form 50 cc. Nessler tubes and the resulting colors are compared after the solutions have stood 40 minutes. 6 Simmonds used 0 5 cc. of the oxalic acid solution but this amount was found insufiicient t o reduce readily the excess permanganate when the proportion of ethyl alcohol was reduced t o 0.5 per cent, and hence 0.7 cc. was used in each case in order to make them exactly comparable. That this slight difference in the amount of oxalic acid used was not an important factor was proven by making a similar comparison in which the two cases differed only in this respect, when there was no appreciable difference in the 8
result. This strength oxalic acid is used by Simmonds but it was found that a 10 per cent solution (10 g. t o 100 cc.) answers the purpose just as well and for the sake of simplicity appears preferable. Hence in all the subsequent work a 10 per cent oxalic acid solution was employed. In cool weather this solution may need a little warming to redissolve the crystals formed on standing.
2 96
T H E J O C 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
of NO. I having increased while KO. 2 still remained almost completely colorless in comparison. According to Simmonds, with “properly sensitive” Schiff’s reagent 0.3 mg. of methyl alcohol in j cc. can be detected but t h e best depths of color for comparison are obtained when the 5 cc. taken for the test contain from I to 4 mg. The standards for t h e colorimetric comparisons are made t o vary by whole units of a milligram, containing I , 2 , 3, etc., mg. of methyl alcohol, respectively. By reducing the proportion of ethyl alcohol from I O t o 0 . 5 per cent, however, and applying the test as outlined above, even as little as 0.1 mg. of methyl alcohol in 5 cc. showed a coloration on standing 40 minutes.’ A suitable series of standards for the colorimetric comparisons may, therefore, be obtained by varying the amounts of methyl alcohol in fractions of a milligram as 0 , 0.1,0.2, 0.3, 0.4, 0.5, 0.6 and 0.7 mg., respectively. PREPARATION OF T H E REAGENT
As has already been indicated above, Simmonds refers t o a “properly sensitive” Schiff’s reagent. He does not, however, say anything descriptive about it, either as t o its composition, mode of preparation or keeping qualities. According t o Schaffer,2 the reagent should be prepared from fuchsine and sulfur dioxide, using the proportion of 0.5 g. of the former t o I g. of the latter in 400 cc. When more than I g. of SOz was added t o 0.5 g. of fuchsine, Schaffer found t h e reagent valueless after standing z days; b u t when prepared in the above proportion, he found it t o keep well for I O days, although he recommends not t o use a solution over 7 days’ old. In this connection, i t was thought t h a t probably an improvement in t h e preparation of this reagent could be introduced b y substituting an equivalent amount of a n h y d r o u s sodium sulfite for the sulfur dioxide. The advantage in such a change is apparent, since it enables one t o weigh out directly the amount required whereas with the sulfur dioxide it probably would be necessary t o prepare first an aqueous solution of it, determine its strength, and then calculate how much of it would be required in order t o have the proper proportion of SOz t o fuchsine. As a matter of fact, Denighs in his communication3 refers t o the variety of procedures which have been given for t h e preparation of Schiff’s reagent and then states t h e one he adopted which calls for a certain volume of a sodium bisulfite solution of a certain degree BaumC. Since, however, sulfite in solution is not very stable while the a n h y d r o u s sodium sulfite is quite stable4 even when kept under ordinary conditions, it appears preferable to base the formula for the fuchsine bisulfite (Schiff’s) reagent on the latter rather than on the former. The following reagent was found t o give satisfactory re1 In working with these smaller quantities of methyl alcohol, the color develops more slowly, so that while 40 minutes was sufficient to show a coloration in comparison with the control even when dealing with only 0.1 mg. in 5 cc., an hour or even two hours should be allowed for the color to develop when the problem is the detection of methyl alcohol, since under these circumstances i t is, of course, desirable to obtain a fairly intense coloration. * U.S. Naval Medical Bull., 6 (19121, 392. 8 Compl. rend., 160 (1910), 530. 4 Elvove, Am. J . Pharm., 82 (1910), 211.
Vol. 9, No. 3
sults and it appears t o be a t least as good, if not better, in keeping qualities’ than the reagent in the preparation of which sulfur dioxide was used as recommended by Schaffer: 0 . 2 g. finely powdered fuchsine2 dissolved in about 1 2 0 cc. hot water and cooled to room temperature; 2 . 0 g. anhydrous sodium sulfite3 dissolved in about 2 0 cc. water and added to the fuchsine solution; then add 2 . 0 cc. HC1 (sp. gr. 1.19) and dilute to zoo cc. with water. After standing for about an hour, this solution is ready for use as a reagent for formaldehyde or methyl alcohol by t h e procedure referred to above. EFFECT OF TEMPERATURE
Since there is a rise in temperature when the solution is mixed with the concentrated sulfuric acid, i t was thought desirable to determine whether or not this rise in temperature affects t h e color produced on mixing with the fuchsine bisulfite reagent. The following experiment was, therefore, carried out: Two j cc. portions of an aqueous solution of methyl alcohol, each containing 0.3 mg. of methyl alcohol and 0 . 5 per cent of ethyl alcohol, were treated by the permanganate procedure referred t o above.‘ After mixing with the I cc. concentrated sulfuric acid, No. I was allowed t o cool t o room temperature before i t was treated with the 5 cc. of fuchsine bisulfite reagent while No. 2 received t h e fuchsine bisulfite immediately after the solution had been mixed with t h e sulfuric acid. After both mixtures had stood 40 minutes, the resulting colors were compared. It was found t h a t they were not alike b u t t h a t t h e one (No. 2 ) which was treated with the reagent immediately after t h e mixing with the sulfuric acid had a deeper color. It appears, therefore, t h a t when a quantitative result is desired the solutions to be compared should all have the same temperature before the fuchsine bisulfite reagent is added. I n most cases this can probably be more advantageously attained by allowing all the solutions to cool t o room temperature after the final mixing with the concentrated sulfuric acid. EFFECT OF T H E PRESEXCE OF FORMALDEHYDE
According t o Simmonds, when the solution t o be examined for methyl alcohol also contains formaldehyde, “its effect must be determined and allowed for.” From this brief and unexplained statement one might get the impression t h a t all t h a t is necessary is t o determine the effect of the formaldehyde on the reagent previous t o the oxidation of the methyl alcohol t o formaldehyde and then allow for this b y subtracting the amount of formaldehyde found before the oxidation from t h a t found after the oxidation. Such a procedure, however, would lead t o erroneous results, since it appears t h a t the oxidation of the methyl alcohol t o formaldehyde in the procedure outlined above is not quantitative but t h a t some of the formaldehyde 1 A fuchsine bisulfite reagent prepared as here described, which had stood in a closed bottle for about six weeks, was found at the end of that time to be as useful a reagent for formaldehyde as one which had been freshly prepared. 2 The fuchsine used had the label of G. Grdbler & Co., Leipzig. 8 Commercial anhydrous sodium sulfite which complies with the requirements of the U. S. Pharmacopoeia may be used for this purpose. 4 See details of experiment at beginning of article and also the accompnnying foot-note.
Mar., 1917
T H E J O U R N A L OF I N 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
is still further oxidized and does not show up in the final reaction with the fuchsine bisulfite reagent. I n the presence of formaldehyde, therefore, i t appears necessary t o proceed somewhat as follows: Determine the amount of formaldehyde in the solution colorimetrically by means of the fuchsine bisulfite reagent and a suitable series of standards containing known amounts of formaldehyde. Then make up the methyl alcohol standards so t h a t they will also contain formaldehyde in the same concentration as t h a t in the solution t o be examined. After this has been done, the procedure referred t o above for estimating the methyl alcohol may be followed and the figures fm methyl alcohol obtained directly by comparison with these modified standards. HYGIEXIC LABORATORY,
u. s.
%'ASHINGTON,
PUBLIC
D.
c.
297
mained except t o consult actual analytical figures. We were able to obtain the figures of herd milk and of t h a t of a large number of individual cows of known purity in papers by Lythgoel and Sherman* and from our own work. The results of all three of these sources, about 1,600 samples, were separately plotted. Using the fats as ordinates and the solids-not-fat as abscissae, curves of the same general shape but on different portions of the chart were obtained. The making of a zone with the extreme curves as boundaries suggested itself t o Mr. Ekroth, giving the interior zone bounded by the lines K - K and P-P as shown in Fig. I herewith. 6.1
HEALTH SERVICE
55 RELATION OF THE FAT IN MILK TO THE SOLIDS-NOT-FAT By LUCIUSP . BROWNAND CLARENCE V. EKROTH
5.0
Received December 1, 1 9 1 6
I n the year 1910the New York State standard for the chemical composition of milk was placed by the Legislature a t its present figures, t o wit: 3 per cent fat and 11.j per cent total solids. No standard for solids-not-fat was mixed. In making municipal standards the City of New York was empowered under the law t o enact additional legislation but could enact no legislation conflicting with t h a t of the State. I n endeavoring t o secure a good milk supply for t h e city and, a t the same time, to make figures which would be useful in the detection of adulteration, the only additional legislation which the city could enact was t o set a standard for solidsnot-fat, which was accordingly placed a t the difference between the State's standards for fat and total solids, namely 8. j per cent. When a rigid enforcement of this standard was attempted it was objected by dealers t h a t the standard was an impossible one and t h a t the cattle in the city's milk shed could not supply it. An investigation was, therefore, set on foot t o determine whether these claims were justified. I n the ccurse of this some very interesting facts developed. I n looking into the experiences of the several states as (presumably) set forth in legislation, it developed t h a t apparently the standards for t h e different states had been made entirely without system. For instance, one state requires a minimum of 1 2 per cent total solids, but only 2.5 per cent of this need be fat. Whether any normal cow could be found in t h a t state (or anywhere else) giving milk containing 9.5 per cent solids-not-fat and 2 . 5 per cent butter-fat is not only open t o a t least a reasonable doubt but i t is quite certain t h a t no unadulterated herd milk would even approach such figures. Another state requires 9.75 per cent solids-not-fat. We venture t o say t h a t a literal enforcement of this standard would leave the state without a milk supply. The legal standards furnishing no help, nothing re-
4.5
4.1
3.5
3.0
2.5
PER
2.0
Ccnlr. Bakt. Parasitcnk.. I I Ab;.. 86, 4 9 1 .
I
8.0
C€NT
SOLJDS N O T
1 6.5
9.0
FAT