the failure of leguminous crops, and the liability of cruciferous crops to certain fungoid diseases. Sour soils redden blue litmus paper when it is pressed on to the moist soil, so that the obvious view was avanced that the infertility is due to the presence of acids, which are neutralised by the application of lime. Methods for the determination of the amount of this acidity and therefore of the "lime requirement" of the soil, were devised on the assumption that the interaction between a sour soil and a base was a simple neutralisation. (Examples of such methods are the Veitch method involving the addition of lime water in slight excess, and the Hutchinson-McLennan method in which the CoCO, removed from a calcium bicarbonate solution is estimated).

A second group of methods depended on the curious fact that sour soils set free considerable amounts of titratable acid from salt solutions such as KCl or KC2H,O2.

A large amount of work has been devoted to the investigation of the validity of these lime requirement methods. Several of them have provided quite useful practical information, but their application is circumscribed by the fact that it is only on comparable soils that the "lime requirements" adequately represent the response to limeing. Further, all the methods are very sensitive to slight variation in technique, and different methods give widely divergent results when applied to the same soil. In a recent study of the Hutchinson-McLennan method, Fisher has shown that the amount of CaCO, removed from solution is not independent of the concentration of the bicarbonate used, but that the distribution of CaCO, beween soil and solution may be represented by the usual adsorption equation. The taking-up of base by a sour soil appears to involve not only the simple neutralisation of acids, but also some absorptive process by the soil colloids.

Many workers have attached more importance to the absorption than to the presence of acids. It has been held that in fertile soils, well provided with reserves of CaCO,, the absorptive power of the soil colloids for bases is fully satisfied, but that the continuous removal of the basic material by plant growth and drainage may result ultimately in a colloidal system with a marked affinity for bases. Ramann therefore replaced the term "acid" as applied to soils by "adsorptively unsaturated." Baumann and Sully, working with peat soils, and Cameron and Harris with mineral soils, denied the existence of acids in soils giving aqueous neutral to litmus paper. Selective adsorption of bases was held to account not only for the action of litmus paper and for the liberation of acid from salt solutions, but for the large variations in the amounts of different bases taken up by the soil from alkaline or neutral solutions. The possibility of establishing the existence of acids in the soil by titration was the subject of prolonged controversy. Truog, for example, maintained that by using diluted solutions to reduce side-reactions, equivalent amounts of NaOH, Ca(OH)2, and Ba(OH), were neutralised by sour soils.

The Measurement of the H-Concentration
of Soils.

The lime requirement methods were intended to measure the "quantity factor" of acidity, i.e., the total amount of acid as measured by titration.

But the acid properties of a solution can only be expressed by the "intensity factor," i.e., the hydrogen ion concentration (or activity). Many sour soils have been shown to possess a H-concentration considerably greater than that of water, so that there is now no valid reason for denying the existence of acid substances in sour soils.

The use of the hydrogen electrode is practicable in complex mixtures only if they possess considerable buffer action, i.e., some mechanism for maintaining an almost unchanged reaction on the addition of small quantities of acid or alkali. Aqueous extracts of soils are but feebly buffered; the introduction of a little alkali, such as the ammonia produced by the reduction of nitrates by the hydrogen electrode, causes considerable alteration in reaction. Suspensions of soil in water, however, are strongly buffered, and give quite definite results with the hydrogen electrode. A very wide range of H-concentrations has been recorded; thus Gillespie found values for log [H] or pH from 44 to 86. Sharp and Hoagland with. a somewhat less satisfactory technique found a range of 37 to 93, though the fertile soils examined by them showed very similar values, varying from the neutral point to about 75 on the alkaline side.

In

The potential difference which forms the basis of the method can be measured with a high degree of accuracy, but with soils difficulties are introduced by the necessity of working with suspensions. Measurements cannot be made on soils at their natural water or CO2 contents. well-buffered mixtures, dilution produces little alteration in reaction, and it has been found that variation in the soil-water ratios or the addition of moderate quantities of CO, to the hydrogen used, produces but little alteration in the results obtained. It is therefore generally assumed that the values obtained aproximate to those of the soil solution.

The sulphone phthalein indicators of Clark and Luhs have been used for the determination of the H -concentration of aqueous soil extracts, and, in the only comparison hitherto made with the results for soil suspensions by the hydrogen electrode, Gillespie has obtained very satisfactory agreement. This is especially important because the most probable sources of error in the two methods arise from different causes. The principal difficulty involved is the preparation of sufficiently clear soil extracts for colour matching. Prolonged centrifuging is often insufficient and it is then necessary to resort to dilution, since filtration or flocculation often produces quite marked changes in such weakly buffered solutions.

Considerable importance has been attached in the past to the differences towards litmus paper exhibited by the moist soil and the soil extract. An insensitive indicator such as litmus, especially on paper, requires an appreciable amount of acid to effect the change to red. With small quantities of weakly buffered solutions, such as soil extracts or very dilute HCl, the reaction of the liquid may be seriously changed, whilst if a large bulk of solution be left in contact with the litmus, leaching may occur before the correct tint is reached. When the paper is pressed on to the moist soil, the proximity of the soil allows rapid diffusion of acid to the paper, whilst its reaction is changing, and a fair indication of the soil reaction may be obtained.

The mechanism of buffer action in the case of partially neutralised weak acids is easily explained in terms of the dissonation constant, but it is not generally possible to analyse the factors involved in more complex systems. The difficulties of technique have prevented much work on the determination of the changes in -log [H] during the titration of acid soils with bases, but it has been found that the soils are wellbuffered over considerable ranges of H⚫-concentration. In other words, the titration curves showing -log [H] as a function of the amount of base added are practically linear without any sharp breaks corresponding to "end points." The slopes of the curves vary considerably from soil to soil, so that it is possible for a strongly acid soil to possess little titratable acidity or a low lime requirement, whilst neutral or even slightly alkaline soils may show an appreciable requirement if the soil be brought into equilibrium with an alkaline solution.

Action of Neutral Salts on Acid Soils. Sharp and Hoagland have shown that BaCl2, KCl, or NaCl increase the H-concentration of sus pensions of acid soils. In a more detailed study of the action of salts on humus suspensions Gillespie found that the same action occurs and that the increases produced by different chlorides are in the order Ba>Na>K. The interpretation of these results is complicated by the fact that these salts produce similar, though smaller, increases in the H-concentration (or activity) of true solutions.

The increase in H-concentration is accompanied by considerable amounts of titratable acid, the estimation of which was used in some of the early lime requirement methods. In spite of much investigation, the interpretation of this action is still obscure. On the selective adsorption theory the action is regarded as strictly analogous to the the liberation of acid and fixing of base during the precipitation of colloidal solutions of As,S, by BaCl. The importance attached to the adsorption of cations for the coagulation of negative colloids gave some support to this view, but no clear mechanism was advanced to explain the liberation of acid.

the differences in the hydrolysis and solubilities of the various salts concerned will account for many of the results. The excess of a salt of a weak acid prevents any considerable rise in H -concentration, so that the aluminium salts are almost completely hydrolysed and precipitated with the liberation of much-titratable acid, whilst with salts of strong acids higher H-concentrations occur, so that the Al remains in solution and the reaction soon reaches an equilibrium point. Similar reactions of the aluminates and silicates of Ca or K will account for the greater adsorption of Ca during the action of the hydroxides on sour soils. In the case of peat soils Sven Odén suggests that the acidity of salt extracts arises from the displacement of small amounts of organic acids adsorbed by humus. He found no HCl by distillation of a KCl extract of an acid peat, and an aqueous extract of the treated peat was not alkaline, so that there was no evidence of any hydrolytic decomposition by adsorption.

The complexity of the soil system is such that nothing is to be gained at present by formulating any general theory of the origin and nature of soil acidity.

Relation of Soil Acidity to Plant Growth. Until recently soil acidity has been regarded as a pathological condition to be avoided by liming, but the reaction of the soil may vary within fairly wide limits and it does not follow that

log (H)=707 is the optimum for plant growth. Soil fertility is the resultant of the interaction of a large number of factors on the plant, and these factors are likely to be influenced to different degrees by the reaction of the soil solution. There is evidence that a distinctly acid reaction may be desirable for certain crops, and attempts are being made in New Jersey to increase the acidity of potato soils by the production of H2SO, through the oxidation of sulphur added to the soil.

The reaction of acid soils is frequently of the same order as that of plant roots, so that the infertility of most acid soils would not appear to be due to the direct action of the acid on the roots. Of the indirect effects Truog attaches the greatest importance to the lack of available basic material, such as CaCO,, for the neutralisation of the acids produced in the plant, whilst several workers have ascribed the infertility to the toxic action of aluminium compounds. Gillespie has recently made some preliminary measurements of the reduction potentials of water-logged soils, and has suggested that the “sourness" of poorly drained soils may be the result of their high reducing power.

An extensive bibliography of soil acidity is given by E. A. Fisher, Journal of Agric. Science, 1921, xi., 19.

MISSISSIPPI RIVER.

By RUSSELL MORGAN.

THE water supply of Clinton, Iowa, situated on the Mississippi River, comes from a series of five wells located above the city. The wells are 200ft. to 250ft. from the river, and from 1800ft. to 2200ft. in depth. The numbers here given express the amounts of the various substances in a million parts of the water.

The cloudiness of the unfiltered river water seems to be due to a small quantity of silica and calcium carbonate in suspension.

The river water and well water contain a relatively large amount of chlorides, and the waters in most other respects are quite similar. Both are rather soft waters for this section of the country, where there is an abundance of dolomite rock. Although the wells are encased, there is a possible connection between the river and the wells.

Cornell College, Mount Vernon, Iowa.

June 4, 1921.

AN EXTRAORDINARY

NUMERICAL RELATIONSHIP BETWEEN CALCIUM AND STRONTIUM.

By ALEXANDER SAKOSCHANSKY, B.Sc.

IT has been frequently stated that atomic weight numbers are something more than relative numbers. While with the present system numerical relationship of a simple type exists only with a few elements, such as √2:43=15588, 232-529, 39122-1529, a system based on other numbers would place other elements in a numerical relationship.

The following is of particular interest, because the values have been derived in a systematic manner. If 011063 (log10 1290) be used as a multiplicand for atomic weights, it will give several pairs of corresponding values, e.g.,

Selenium X 11063 = Strontium
Sulphur 11063 = Chlorine.

Now log1018784=0*27379

and log103 5284=0*27379

2

or 3528418784

Barium shows an interesting numerical relationship which cannot be described in this short article.

In addition it should be noted that on the basis Cl=25 46, fluorine and iodine are in logarithmic decrement, for F = 13'64, I=91'12, 10×1/F = logx I, r = anti-logo 1365=1.352. Magnesium 17°46.

A STUDY OF ARSENICAL COMPOUNDS RELATED TO ARSPHENAMINE.

By GEORGE W. RAIZISS and JOSEPH L. GAVRON.

Introduction.

EXPERIENCES of the last decade have shown to the leading syphilologists that arsphenamine, which is the official American name for 3,3'-diamino4,4'-dihydroxy-arsenobenzene dihydrochloride (P. Ehrlich and A. Bertheim, Ber., 1912, xlv., 756), is one of the most efficient remedies in the treatment of syphilis. It has been observed by many that intravenous injections of this drug sometimes produce alarming symptoms. While these have been attributed to various causes, some relating to the condition of the patient, and others relating to he technic of adminisration, it seemed very probable to us that chemical impurities in arsphenamine may produce the disturbing effects. Therefore, we deemed it advisable to prepare a number of compounds, which are closely allied to arsphenamine and which might appear as impurities in the drug in the course of manufacture.

tetra

Arsphenamine is produced by the reduction of 3-nitro-4-hydroxy-phenylarsinic acid which some times contains the corresponding dinitro derivative as an impurity. It is possible, therefore, that arsphenamine may be contaminated with 3,5,3',5'tetra-amino-4,4'-dihydroxy-arseno-benzene hydrochloride. In view of this fact, it was of interest to prepare the latter compound and study its chemical and biological properties (J. F. Schamberg, G. W. Raiziss. J. A. Kolmer, and J. L. Gavron "The Chemotherapy of Arsenical Compounds Related to Arsphenamine" (to be published

This compound has been very briefly mentioned in a German patent (D. R. P., 224,953, 1910), in which the method of preparation consists of the reduction of 3,5-dinitro-4-hydroxy-phenyl-arsinic acid by means of sodium hydrosulphite. same compound has been prepared by us according to the following procedure.

The

Twenty g. of 3,5-dinitro-4-hydroxy-phenylarsinic acid is suspended in 50 cc. of water and dissolved by adding 13 cc. of 40 per cent sodium hydroxide (2 mols.). To this solution is added a solution of 40 g. of magnesium chloride (crystals) in 400 cc. of water, the whole cooled down to o° and reduction effected by gradually introducing 130 g. of sodium hydrosulphite (commercial powder) with continuous mechanical stirring. During this process, the temperature should not be permitted to rise above 2°, since otherwise further reduction to the arseno compound will occur. When all the hydrosulphite has been introduced, the solution becomes decolorised, and upon continued stirring for 15 hours at o° a white crystalline precipitate of the amino compound appears. This is filtered off and the mother liquor expressed as completely as possible. To purify, the crude product is dissolved by triturating with an excess of diluted hydrochloric acid in a mortar. The solution is filtered, and the filtrate neutralised at low temperature with concentrated sodium hydroxide solution until it is but slightly acid to congo red paper. The precipitate of the pure amino compound is filtered off and washed with cold water until it is free from sulphate and chloride, then with methyl alcohol, and finally with ether. is dried in vacuo over sulphuric acid. The yield (after one purification) is about 80 g. or 50 per

cent.

:

It

Properties. 3,5-diamino-4-hydroxy-phenylarsinic acid is a white crystalline product which darkens during the process of purification. It is soluble in alkalies, also hydrochloric acid insoluble in methyl alcohol, ethyl alcohol, ether, acetone, chloroform, benzene. Its solution in alkali darkens very rapidly on standing exposed to the air.

Calculated for C.H,O,N2As: N, 11.29; As, 30.24. Found: N, 10.70; As, 29′52.

3-Amino-4-hydroxy-phenylarsinic Acid. This compound was first prepared by Ehrlich and Bertheim by the reduction of 3-nitro-4-hydroxyphenylarsinic acid with sodium amalgam (D.R.P., | 224,953, 1910). Another method in which sodium hydrosulphite is used as the reducing agent has been briefly described in a German patent (D.R.P., 224,953, 1910). Jacobs and Heidelberger Jour. Am. Chem. Soc., 1918, xl., 1580) prepared this same arsinic acid by the use of ferrous sulphate and sodium hydroxide. The authors in preparing this compound used sodium hydrosulphite as the reducing agent.

Twenty g. of 3-nitro-4-hydroxy-phenylarsinic acid is suspended in 50 cc. of water and dissolved by adding 15 cc. of 40 per cent sodium hydroxide solution (2 mols.). A solution of 40 g. of magnesium chloride crystals in 300 cc. of water is added, the whole cooled down to o°, and reduced by adding 85 g. of sodium hydrosulphite and proceeding exactly as in the case of the previous compound. The yield of purified product is about 108 g. or 61 per cent.

Properties.-3-Amino-4-hydroxy-phenylarsinic is a white crystalline substance which during purification turns slightly brown. It is soluble in hydrochloric acid, sodium and ammonium hydroxides; insoluble in acetic acid and the usual organic solvents. Its alkaline solution darkens rapidly on standing in air.

Calculated for CH,O,NAs: N,601. Found 5.74. Acetyl-3-amino-4-hydroxy-phenylarsinic Acid.Ten g. of 3-amino-4-hydroxy-phenylarsinic acid is suspended in 50 cc. of water, immersed in an icebath, and mechanically stirred. Nine cc. of acetic anhydride (one mol equals 44 cc.) is gradually introduced, and the stirring maintained for three hours. The whole is filtered, and the precipitate suspended in dilute hydrochloric acid. This removes the unchanged amino compound, the acetyl derivative remaining undissolved. This is filtered off, washed with water till free from hydrochloric acid, then with methyl alcohol, and finally ether. It is dried in vacuo over sulphuric acid. The yield is 35 g. of pure substance.

Properties.-Acetyl-3-amino-4-hydroxy-phenylarsinic acid is a light brown crystalline substance, soluble in dilute alkalies. It is insoluble in hydrochloric acid, acetic acid and the usual organic solvents.

Calculated for C,H,,O,NAs: N, 5'09; As, 27.27 Found: N, 5'10; As, 27.61. Diacetyl-3,5-Diamino-4-hydroxy-phenylarsinic Acid.-This compound is prepared from 3,5diamino-4-hydroxy-phenylarsinic acid in exactly the same way as the previously described compound. The yield is 75 per cent.

Properties. A pale brown crystalline substance, soluble in sodium and ammonium hydroxides. Insoluble in water, hydrochloric acid, and the usual organic solvents. Calculated for CH,,N,O,As: N, 8'43; As, 22.60 Found: N, 849; As, 22.98 per cent. 3,5,3',5'-Tetra-amino-4,4'-dihydroxy-arsenobenzene Tetrahydrochloride. Although this amino compound has been very briefly described in a German patent, its tetrahydrochloride has not been described. It is interesting to note that it appears to exist in two modifications, one easily soluble in methyl alcohol, and the other only sparingly so.

25 g of the corresponding dinitro-hydroxyphenylarsinic acid is dissolved in 525 cc. of water to which has been added 16 cc. of 40 per cent sodium hydroxide solution (2 mols.). Reduction is effected by a solution of 575 g. of sodium hydrosulphite and 65 g. of magnesium chloride in 2600 cc. of water. After 15 hours' stirring at 55° to 60° the yellow precipitate of the tetra-amino compound is filtered, washed with distilled water until most of the hydrosulphite has been removed, and then dried on porous plates.

To convert the dried base into its tetrahydrochloride, it is introduced into 220 cc. of absolute

The

methyl alcohol containing hydrogen chloride. About one-half of the tetrahydrochloride dissolves. This is decanted off. The remaining insoluble portion is triturated in a mortar with more ethyl alcohol-hydrochloric acid solution in order to remove the soluble fraction completely. entire methyl alcohol solution of the tetrahydrochloride is mixed with five volumes of U. S. P. ether. A light yellow precipitate is immediately produced. This is allowed to settle, the supernatant fluid decanted, the precipitate washed several times by decantation with ether, rapidly filtered and finally dried in vacuo over sulphuric acid. The yield is 80 g.

The undissolved fraction of the tetrahydrochloride is washed with ether and also dried in vacuo over sulphuric acid. The yield is 75 g.

Both modifications are yellow substances easily soluble in cold water. From this solution the base is precipitated by the addition of sodium hydroxide and redissolved when sufficient alkali has been added to form the monosodium salt. This solution on standing rapidly becomes turbid due to the action of the carbon dioxide of the air. A solution of disodium salt is less readily precipitated. Both salts in solution when exposed to the air, rapidly oxidise, more so than the corresponding salts of 3,3'-diamino-4,4'-dihydroxyarsenobenzene dihydrochloride.

Analysis. Neutralisation of the hydrochloride and formation of monosodium salt. Soluble modification: subs., 0100: 90 cc. 01 N NaOH Insoluble modification: subs., O' 100: 75 CC. O'I N NaOH. Calculated for C,,H,,N,CÍO,As,. ICH, OH 87 cc. o'1 N NaOH. Oxidation by iodine. Soluble modification. 01 N iodine.

18

Subs., 0100: 13'50 CC.

Insoluble modification. Subs., 0'100: 13'6 cc. O' N iodine.

Calculated for C12H18N,CO2AS2. 1CH,OH : 13'95 CC. O' I N iodine; 11.16 mg. oxygen Found: Sol. mod., 10.80 mg. oxygen; insol. mod., 10.88.

In view of the fact that this compound cannot be recrystallised, the only method of purification consisting of redissolving in methyl alcohol-hydrochloric acid solution and reprecipitating from ether, we are reporting these results as the best we have been able to obtain thus far.

Calculated for C12HN.CI,O,As2. ICH,OH: N, 976; As, 26 13. Found: sol. mod., N, 8.80; As, 26 62; insol. mod., N 8.80; As, 25'09 per cent.

Tetra-acetyl-3,3′,5,5'-tetra-amino-4,4'-dihydroxyarsenobenzene.-Two g. of diacetyl-3,5-diamino-4hydroxy-phenylarsinic acid is dissolved in 70 cc. of water by the addition of 4 cc. of 15 per cent sodium hydroxide solution. Two g. of magnesium chloride (crystals), and 10 g. of sodium hydrosulphite are dissolved in 45 cc. of water and immediately mixed with the solution of arsinic acid. The combined solutions are filtered rapidly, and warmed up to 55 or 60°. This temperature is maintained for 30 minutes, during which time the mixture is continually stirred. The precipitate which forms is yellow at first, but at the end of the reduction it appears white. It is filtered off, washed thoroughly with water and dried in vacuo over sulphuric acid. The yield is 12 g.

The compound is a white powder insoluble in water, dilute hydrochloric acid, and the usual or

ganic solvents, soluble in glacial-acetic acid however. It is soluble in dilute alkalies, forming a yellow solution which is characteristic of arseno compounds; also soluble in dilute sodium hydrogen carbonate solution. Like all other arseno compounds it decolorises iodine, thereby becoming oxidised to its corresponding arsinic acid. This reaction cannot be utilised for its quantitative estimation, thereby differing from other arseno compounds, because of difficulty in distinguishing the exact end-point.

Calculated for C,H,,O,NAs,: N, 9'93; As, 26.60. Found: N, 1014; As, 25 75.

Diacetyl-3,3'-diamino-4,4'-dihydroxy-arsenobenzene. The method of preparation of this substance from acetyl-3-amino-4-hydroxy-phenylarsinic acid is the same as that of the tetra-acetyl

compound described above. The yield is 73 17

per cent.

THE production of pyrites was, in 1912, 82,600 metric tons, and 126,000 in 1913, since when it has increased. Moreover, we must consider the fact, that Japan is a great sulphur producer which favours manufacture of acid. Before the war, the production was, apparently, 30,000 metric tons, which has increased to 70,000. Subsequently to an agreement existing before the war, which was then cancelled and again restored, prices were fixed by a Convention in 1916. Sulphuric acid, as in all other places, is utilised for manufacture of mineral fertilisers, especially superphosphates. Imports of natural phosphates which, in 1913, amounted to 347,000 metric tons, fell to 322,000 in 1914, and even 189,000 in 1915; but the imports for 1916 and 1917 were estimated at 280,000.

are

However, Japan has developed her phosphate production in the Island of Rasa, Lu Choo, which from 740 metric tons in 1908 rose to 38,259 in 1914; 57,716 in 1915, and 114,810 in 1916. Although the superphosphates obtained largely consumed in the country itself, a considerable quantity is exported to Java, where 40 per cent superphosphates are required, and also to Russia. This market was only of interest during the war. It seems possible to greatly develop the utilisation of artificial fertilisers in Manchuria, where they are as yet little employed.

There are three principal manufacturing_companies of fertilisers, viz., the Kuhard, the Furukawa, and the Sumitomo Co., the respective annual production being 7,000,000, 4,880,000, and 680,000 bags.

Consumption of sulphate of ammonia is about 140,000 metric tons. chiefly supplied by imports from England (110,000 metric tons in 1910). As during the war these imports almost ceased (20,000 in 1915) the Japanese have remarkably developed the industry.

The production of sulphate of ammonia was