ÆäÀÌÁö À̹ÌÁö
PDF
ePub

WORK ON MIXED FERTILIZERS.

Mix thoroughly before sampling.

(1) In sample No. 2 determine water-soluble potash (K2O) by the official method (U. S. Dept. Agr., Bureau of Chemistry Bul. 46 Revised, p. 21, or Bul. 107, p. 11).

(2) Use proposed volumetric method for potash in fertilizers. Transfer 10 cc of solution No. 2 (Bul. 46 Revised, pp. 21-22, or Bul. 107, pp. 11-12) to a platinum dish, and add 0.25 cc of sulphuric acid (1 to 1). Evaporate to dryness and ignite to whiteness. Dissolve residue in hot water plus a few drops of hydrochloric acid and transfer to a tall 200 cc beaker; add 30 cc of phosphomolybdic acid solution and proceed as under proposed volumetric method for potash (K2O) in soils. (If excess of phosphomolybdic acid has been used, the dried residue has a reddish hue. If excess has not been added, the residue is bright yellow. The residue should not appear white.)

In each case run blanks to ascertain corrections to be made for impurities. It is also advisable to treat the potassium-platinic-chlorid residue in the gooch crucible in order to ascertain if it is all soluble in water. Then reweigh the crucible after thoroughly drying.

(See Sut

Note. The caustic potash solution must be free from carbonate. ton, Volumetric Analysis, 9th edition, revised, page 48, for the barium hydrate method, also page 301 for the alcoholic method for making potassium hydroxid free from carbonates.)

A. L. KNISELY, Referee.

B. B. Ross, Associate Referee.

Ten sets of samples were sent out and three reports have been received.

TABLE V.-Comparison of official and proposed volumetric method for the determination of potash in soil and in mixed fertilizers.

[blocks in formation]

M. G. Donk: Could obtain no satisfactory results on sample No. 2 with the volumetric method, presumably due to difficulty in washing yellow precipitate. G. S. Fraps: Sample of phosphomolybdic acid from Merck left considerable white residue, probably molybdic acid, when diluted in nitric acid. This probably accounts for volumetric results being higher than the official method.

C. E. Bradley: Best results were always obtained when using small amounts of potash solution. An aliquot part of a solution of fertilizer containing not more than 0.02 gram K2O gave the best results. If, after evaporating to dryness and taking up in dilute nitric acid, the insoluble residue is not bright yellow, then the conditions are not normal. The washed precipitate should be bright yellow.

COMMENTS BY REFEREE.

It was recommended this year to use a stronger nitric acid wash solution than was formerly used. This was on account of the difficulty experienced in washing the yellow precipitate free from impurities. It has since been ascertained that this difficulty was due to poorly prepared or old phosphomolybdic acid solution. The phosphomolybdic acid solution, when prepared from good, fresh chemicals, should give no blank, and when used in volumetric potash work dissolves and washes out very readily from the evaporated residue. It has recently been found that the nitric acid wash, instead of being stronger than was formerly used, should, if anything, be weaker. In some cases when determining potash in material containing considerable calcium and magnesium and moderate quantities of iron, such as some soil solutions, it is not necessary to use a nitric acid wash solution, but only to dissolve the material at once in the sodium nitrate wash solution.

Official soil solution evaporated to dryness with phosphomolybdic acid and taken up in nitric acid wash gave 0.45 per cent KO; when taken up directly in sodium nitrate wash the same result was obtained.

I think the merit of the volumetric method for potash depends entirely upon the phosphomolybdic acid solution and the care with which the evaporation is conducted. The difficulties thus far experienced with blanks and in washing the evaporated residues have been largely due to the presence of white insoluble molybdic acid, which at times separates during the process of evaporation and is difficult to remove from the yellow residue of potassium phosphomolybdate.

RECOMMENDATION.

The referee recommends a continuation of the study of the volumetric method for use in both soil and fertilizer analysis; also that the recommendation of Mr. Cushman concerning a study of what really constitutes available potash in soils, fertilizers, and ground-mineral products be continued, so that if possible we may have a definition of available potash.

REPORT ON INSECTICIDES.

By R. J. DAVIDSON, Referee.

The suggestions made in the recommendations of last year were followed as closely as possible, and after consultation with the associate referee the following work was outlined:

(1) Work on London purple to test the Davidson modification of Haywood's method in comparison with the method as originally outlined.

(2) Tests to determine which of the two methods of analysis of soda lye give the more correct results.

(3) Testing the formaldehyde method as modified by Haywood and Smith, in comparison with the method as it originally stood.

(4) Testing the Avery method of determining sulphur in sulphur dips.

(5) Testing the methods of examining lead arsenate as published in last year's proceedings.

Requests for cooperation were sent to thirty chemists. Of these, but two signified their willingness to undertake the work. One report contained results obtained by two analysts in the same laboratory. The associate referee was prevented from doing any work on account of a serious injury he received as the result of an accident in his laboratory. The report therefore contains the results of five analysts.

LONDON PURPLE.

The work on London purple consisted in testing the modified Haywood method (for the removal of part of the coloring matter) in comparison with the method as originally outlined. The modification was proposed in 1903 and was given a partial trial in 1904, but no definite results were obtained. A statement was made that it greatly facilitated end point readings. The benefits of this modification are twofold: Time is saved and, by removing a part of the coloring matter, the end point reading in the titration is made less difficult. If, therefore, the modification does not interfere with its accuracy, it will be of great advantage.

In presenting the work done it has been deemed advisable to give the results for arsenious and arsenic oxids in separate tables, as the work on these two determinations has not been equally satisfactory. The subject will therefore be discussed under two heads.

TOTAL ARSENIOUS OXID.

Method I is a provisional method, and may be found in Circular No. 10, Revised, Bureau of Chemistry.

Method II is the proposed modified method, and differs from Method I in that the process of filtering and washing to a definite volume after dissolving the material is eliminated, and in addition a part of the coloring matter is precipitated. The method is as follows:

Place 2 grams of London purple in a beaker and dissolve in about 80 cc of water and 20 cc of concentrated hydrochloric acid at a temperature of 60° to 70° C.; cool, and add sodium carbonate in slight excess, transfer to a 250-cc flask, bring to mark, shake, and filter through a dry filter. Acidify 50 cc of the filtrate with hydrochloric acid and make alkaline with sodium bicarbonate. Titrate the amount of arsenious oxid present with standard iodin solution.

[blocks in formation]

The results upon total arsenious oxid agree very well. There is a difference of only 0.33 per cent between the averages by the two methods, showing that

in case of the total arsenious oxid, the shorter method gives practically the same results.

TOTAL ARSENIC OXID.

Methods I and II are the same as were used in 1904 and 1905, and may be found in Circular No. 10, Revised, of the Bureau of Chemistry. Method II is a modification of Method I, the difference being that before proceeding with the reduction of the solution a part of the coloring matter is removed by precipitation with sodium carbonate. Method III is a still further modification of Method I, as will be seen by reference to Method II under total arsenious oxid. Method III reads as follows:

Acidify 50 cc of the solution prepared under determination of total arsenious oxid, Method II, with concentrated hydrochloric acid, heat to 80° C., add 50 cc more concentrated hydrochloric acid and 3 grams of potassium iodid, and proceed as described in Method I for total arsenic oxid.

The only difference between Methods II and III is the elimination of the filtering and washing to definite volume of the hydrochloric acid solution, which saves considerable time

[blocks in formation]

The results upon total arsenic oxid are very unsatisfactory. There is a fairly good agreement of the results of four analysts under each method, but the results of the different methods do not agree, there being a difference of about 2 per cent between Method I and Methods II and III. One set of results shows an excellent agreement by all three methods. All of the results agree very well by Method I, but by Methods II and III one set of results is about 2 per cent higher than the other four.

It is very difficult, if not impossible, to explain why four sets of results should agree so well among themselves by Methods II and III, yet differ so widely from the other set. If there had been agreement between all the analysts, then the difference between Method I and Methods II and III might' be attributed to one of two things: an error in Method I due to the difficulty of reading accurately the end point in the titration when so much coloring matter is present, or to the loss of arsenic occurring in removing the coloring matter. It was suggested that the difference was partly due to not carrying out properly the details of Method III, as it was thought that the directions for Method III were not explicit enough as to heating the solution before adding

hydrochloric acid and using only 25 cc instead of 50 cc. This can not be the explanation, for in my own work the solution was heated to 80° C. and 50 cc concentrated hydrochloric acid tried as well as 25 cc, giving practically the same results. It is regretted that no comments are made by the analysts on this point, and also that the referee did not have time to investigate it further. Method II was used in 1903 and 1904, but no mention was made of any difficulty occurring; in fact, the referee stated in his discussion that "the modification suggested for the removal of a part of the coloring matter works very well, greatly facilitating the end point reading." In applying this method to the samples of 1903, I did not notice such a difference in results. This part of the work on London purple must be repeated next year, and it is to be hoped that the cause of the discrepancy will be discovered.

SODA LYE.

The methods used for the examination of soda lye were the same as have been used for several years, and may be found in Circular No. 10, Revised, Bureau of Chemistry.

In Method I the total alkalinity is obtained by titrating an aliquot of the dilute solution with half-normal acid, using methyl orange as an indicator. Another aliquot is taken, and the carbon dioxid of the carbonate is precipitated by a dilute solution of barium chlorid and then made to a definite volume. The precipitate is allowed to settle, and an aliquot of this is taken and titrated with half-normal acid, using phenolphthalein as an indicator. In Method II an aliquot is titrated with acid potassium sulphate, using phenolphthalein and methyl orange as indicators.

[blocks in formation]

The results are not so good as they have been in previous years. Two sets of results are slightly lower in sodium hydroxid and higher in sodium carbonate than the others. The results for sodium hydroxid by the two methods agree very well. The sodium carbonate by Method II is higher. This higher result may be due to the difficulty in reading the end point with methyl orange. In Method II the titration is tedious, and for this reason, together with the difficulty in reading the end point in the titration, the referee thinks it should be dropped from the methods. In Method I, when titrating with phenophthalein to determine the sodium hydroxid after precipitating with barium chlorid, it is not necessary to filter off the precipitate, as has been pointed out before, since the precipitate does not interfere with the titration. The results obtained when the precipitate is permitted to remain in the solution are the same as when it has been removed. Better results would probably be obtained in this work if a fifth normal solution of hydrochloric acid were used instead of half normal. A

« ÀÌÀü°è¼Ó »