e

CHEMICAL NEWS,
Aug. 7, 1914

THE

THE

1914
Determination of Lime Requirements of the Soil.

DETERMINATION OF THE LIME REQUIREMENTS OF THE SOIL. By H. B. HUTCHINSON and K. MACLENNAN, Rothamsted Experimental Station.

THE use of lime as a means of soil amelioration has been recognised by agriculturists since the time of the early Romans. Its application as marl, shells, chalk, limestone, and burnt lime is recorded in all the earlier books on British agriculture, and, in fact, the practice persisted until well into the nineteenth century. With the incidence of low prices, bad seasons, and an increasing appreciation of the value of the more quickly acting artificial manures, the practice fell into abeyance some thirty years ago, and since that time not only British but also some of the Continental agriculturists have been living on the heritage left | by their predecessors.

Within this period, however, and with the growth of agricultural science, it has become increasingly apparent from physical, chemical, and biological considerations that a resumption of the practice was, in many cases, of paramount importance if the fertility of the soil was to be maintained. Notably on the Continent, where large tracts of peaty moorland are to be found, the primary importance of lime as an agent in the reclamation of virgin soils was constantly demonstrated, whilst conditions in the eastern states of America were equally indicative of lime starvation. It is natural, therefore, that repeated attempts should have been made to determine the lime requirements of the soil, and of these two chief lines of inquiry may be mentioned. The first consists of the application of lime in various amounts and in different forms to field plots under controlled conditions, whereas the second is represented by the introduction of chemical and other methods to indicate the requirements of any given soil. Whilst attention in Britain has been almost entirely confined to the former type of investigation, much work has been done in Germany and the States in the latter direction, and in Denmark in recent years the results of laboratory methods have been advantageously correlated with conditions actually obtaining in the field.

Although the maintenance of a favourable soil reaction and the neutralisation of the acids formed by biological processes in the soil is secured by means of a solution of calcium carbonate in the soil water, the methods proposed for the determination of the lime requirement of acidity of the soil have rested on a great diversity of chemical reactions, qualitative and quantitative.

The obvious but rough test for soil reaction is the commonly adopted litmus-paper test, and this has been extended by Christensen and Larsen to litmus solution, varying tints produced being taken as a rough measure of the acidity (Tidskrift for Landbrugets Planteavl, 1910, xvii., 407, and Centr. Bakt. Par., [II.], 1911, xxix., 358). Loew advocates the use of a solution of potassium iodide and starch as being roughly quantitative (Zeit. für Landw. Vers. Wesen in Esterreich., 1909, 462).

Numerous methods based on the determination of calcium as carbonate or other salts soluble in dilute acids or salt solutions have been suggested. The importance of a definite amount of calcium carbonate has been strongly urged by English agricultural writers on the assumption that the only effective compounds for soil purposes is the carbonate, and the minimum content has often been placed at o2 to 0.5 per cent. There is no doubt, however, that this is too high, but comparison with Continental standards is rendered difficult by the lack of uniformity in

61

the methods of determination; although it has been frequently shown that treatment of the soil with boiling acids gives rise to the production of carbon dioxide other than that derived from carbonates, this method is still largely practised.

As distinct from these may be mentioned the estimation of lime soluble in various organic and inorganic acids and neutral salts. Kellner suggested a method (Landw. Vers. Stat. 1887, xxxiii., 359), afterwards largely adopted by Meyer (Landw. Jahrb., 1900, xxix., 913), in which the soil is exposed to the action of a solution of ammonium chloride. While the data thus obtained have been of distinct use in many cases there have been others in which the results were directly contradictory, and in further extensive experiments Meyer expresses the opinion that the determination of the acidity of the soil must receive much more attention than has hitherto been the case (Landw. Jahrb., 1910, xxix., Erganz. Bd. III., 287. Christensen and Larsen (loc. cit.) also found by correlation of the results obtained from the ammonium chloride method with those in the field, that with 49 out of 116 soils little guidance could thus be acquired.

Holleman proposed extraction of the soil with water saturated with carbon dioxide and estimation of the lime removed (Landw. Vers. Stat., 1892, xli., 38). Sulphuric acid (Immendorf, Fifth Intern. Cong. App. Chem., Berlin, 1913, III., 740), acetic acid, dilute and concentrated hydrochloric acid have also been used, and have given rise to the creation of standards often without value.

Whereas the above are based on the presence of certain lime compounds, the determination of the lack of bases or soil acidity has been more or less followed. Tacke introduced a method depending on the amount of carbon dioxide evolved ftom an excess of finely precipitated calcium carbonate (Chem. Zeit., 1897, xxii., 174), while in Süchting's modification of this a definite amount of chalk is added (Zeit. Angew. Chem., 1908, xxiii., 151), and the residual carbonate existing after a certain period is estimated.

The capacity of soil acids to displace mineral acids from neutral salts has been employed by Hopkins, Knox, and Pettit by means of potassium nitrate solution (Proc. 19th Ann. Conv. Off. Agric. Chem., Washington, 1902). The reaction was found to be extremely tardy, and the results obtained by successive portions of solution were in decreasing geometrical progression, and for the interpretation of the results a factor is used. Loew (Bull. 13, Porto Rico Agric. Exp. Stat., 1913, 7), and also Jones (American Fertiliser, 1913, xxxix., 29), employ a solution of a neutral acetate for the same purpose, but from results obtained by us it seems apparent that this method is open to the same objection as the preceding one. Gregoire, Hendrick, Carpiaux, and Germain (Ann. Stat. Agron. Gembloux, 1913, ii., 87) have recently investigated the action of soils on "Kjeldahl's solution" consisting of potassium iodide, iodate, and sodium thiosulphate, and the examination of a very large number of soils has permitted of differentiation according to intensity of action, although neutral soils were also found to give a distinctly positive reaction.

Methods involving the use of alkaline solutions such as ammonia have been used by Müntz (Frémy's "Encyclopédie Chimique," iv., 182), and Wheeler, Hartwell, and Sargent (Fourn. Am. Chem. Soc., 1900, xxii., 153), barium hydroxide solution was employed by Albert (Zeit. Angew. Chem., 1909, xxii., 533), and Lyon and Bizzell (Fourn. Ind. and Eng. Chem., 1913, v., 12), whilst calcium hydroxide was used by Wheeler and his colleagues and by Veitch (Fourn. Am. Chem. Soc., 1902, xxiv., 1120).

The method proposed by the latter has found fairly wide adoption in the United States, and gives, as far as our own experience goes, results closely upproximating to the actual. It is, however, a very tedious process, and is difficult of adoption on a large scale.

In the course of other work on the action of lime on the

fertility of the soil, which has been carried out in this | laboratory, the need was felt of a simple and accurate method for the determination of the lime requirement of the soil. Attention has already been called to the general view that carbonates and bicarbonates are the chief compounds tending to maintain a neutral reaction of the soil, and to these certain easily hydrolysable calcium and other silicates may be added. It appeared, therefore, that a closer investigation of the action of certain carbonates on the soil might give a measure of prevailing acidity, and would possibly conform more closely to natural conditions than some of the compounds hitherto employed.

respectively), and the further addition of chalk gives returns varying only slightly from those of the control soils; Millbrook soil, which is almost neutral, gives a slight increase in nitrates and a depression in plant growth. The two acid soils, Woburn and Craibstone, respond readily to chalk applications, as indicated by increased nitrate production and plant growth of the first four crops after treatment.

During the investigations mentioned above, other portions of the same soils were treated with increasing amounts of calcium oxide, and allowed to stand in a moist condition for about twelve months. These have now been examined by this method in order to ascertain how far the original reaction had been affected by the application of such doses. The results are given in Table II., where the acidity is given in terms of calcium carbonate required to neutralise the soil in each case.

Preliminary work with sodium carbonate and bicarbonate gave unsatisfactory results, inasmuch as the deflocculation of the clay compounds of the soil occurred, and adversely affected the rate of filtration; a coloured extract difficult of titration was also obtained, and, furthermore, these compounds were found to give positive absorption even in the case of neutral soils containing abundance of calcium carbonate initially. Recourse was then had to the use of a solution of calcium bicarbonate, and after minor modifications, as to period of digestion, strength of solution, (stated as CaCO Rothamsted. Millbrook. &c., this method has been used with a large number of soils under well controlled conditions.

The solution may be prepared by pass.ng a current of carbon dioxide into a suspension of calcium carbonate in distilled water, or by means of a "Sparklet" or refillable soda-water syphon, where bulbs of compressed carbon dioxide are used. The latter method is the more convenient, and permits of the preparation of a saturated solution within quite a short time. A large excess of carbonate must be used in order to provide an abundance of small particles which readily pass into solution; the contents of the syphon may be diluted with one-third its volume of distilled water before filtering, and this will result in the formation of a solution of approximately N/50 strength.

For a determination of acidity, or lime requirement, 10-20 grms. of the soil are placed in a bottle of 500-1000 cc. capacity together with 200-300 cc. of the approximately N/50 solution of calcium bicarbonate, and the air in the bottle is displaced by a current of carbon dioxide in order to insure against possible precipitation of the calcium carbonate during the period of determination. The bottle is then placed in a shaking machine for three hours, after which time it is opened, the liquid is filtered, and a portion of the filtrate equal to half of the original amount of bicarbonate solution is titrated against N/10 acid, using methyl orange as indicator. The difference between this final titration and that of the initial solution represents the amount of calcium carbonate absorbed, each cubic centimetre of N/10 acid being equal to 5 mgrms. calcium carbonate.

This method has been tested within the last few months on a number of different soils, the behaviour of which has been ascertained bacteriologically and chemically in the laboratory. A few of these results are summarised in Table I., in which the production of ammonia and nitrates and plant growth in untreated and limed soils is given. TABLE I.-The Relation between Soil Acidity, the Production of Ammonia and Nitrates, and Plant Growth

in certain Soils.

Soil. Rothamsted Chelsea

Increase in ammonia
and nitrates (pts. per
mill. dry soil).

Plant growth
(production of dry
matter in 4 crops).

Unt.

Soil.

CaCOs.

soil.

Soil.

CaCO.

105

99

97 740*

144

The two soils, Rothamsted and Chelsea, contained initially an abundance of carbonate (2.6 and 0.8 per cent

TABLE II.-The Acidity of Certain Soils after Treatment

CaO applied

per cent).

0.18

0.36
0'54
0.72

with Lime.

Further pot culture experiments have been made with an acid soil to which definite amounts of calcium carbonate were supplied, and the crop results obtained show quite unmistakably that the acidity indicated by the above method coincides with the physiological gauge set by the plant. The crop growth was found to increase directly with increasing applications of carbonate until the neutral point was reached, and then, with increasing carbonate, remained constant.

In addition to its value for practical agricultural work, the method will possibly be of use in various ecological problems, where the relations between plant and soil require more accurate determination.

OSMIUM-PLATINUM-A NEW ALLOY.*

By F. ZIMMERMANN.

OF the several metals of the platinum group platinum, palladium, iridium, and rhodium have been most generally employed in the industrial arts, either alone or in com bination as bivalent alloys. Of the latter, iridium-platinum is the best known, but the growing scarcity of iridium has led to the search for other combinations of the metals of this group yielding alloys possessing physical and chemical properties of equal if not greater value. The rarer metals of the platinum group are not easily obtained in great purity, and because of this fact but little success has heretofore been obtained when combining them as bivalent alloys. Furthermore, the strong affinity of osmium for with other metals in definite proportions. After much oxygen has increased the difficulty of making alloys of it experimentation by the author, highly refined platinum and osmium have been successfully combined in widely varying proportions yielding alloys of commercial value. While the two metals may be combined in almost any proportion, alloys containing from 1 to 10 per cent of osmium and 99 to 90 per cent of platinum are chiefly used.

Great purity of the components is essential, as the presence of small percentages of other elements appears to be very detrimental to the properties of the resulting alloy. According to the chemical and physical behaviour, it seems that one part of osmium in an alloy with platinum will take the place of two and one-half times its weight of

Paper presented at the Twenty fourth General Meeting of the American Electrochemical Society, at Denver, Colorado.

iridium. The osmium-platinum alloy is very acid-resisting, ethanes. How this decided distortion could be accounted and for this reason may be of great service in the electrochemical industry. Its electrical resistance is considerably higher than that of an iridium-platinum alloy of the same percentage composition. The alloy further possesses great hardness and tensile strength. Wires of the finest size are drawn with comparative ease.

A thorough study of the alloy is in progress, and definite results of tests may be of sufficient interest for a later paper.

The alloy has been patented-U.S. Patent No. 1,055, 199, March 4, 1913.-Chemical Engineer, xviii., No. 3.

THE CHEMICAL SIGNIFICANCE OF
CRYSTALLINE FORM.*

By THEODORE W. RICHARDS.
(Concluded from p. 53).

TURNING now once more to the compound containing I atom of hydrogen, we see here great difference in crystal form. This is also just what we should expect in a compressible system, provided that the newly introduced member of the system possessed not only a different molecular volume, but also a different chemical nature. Under such circumstances a marked change in the habit and a twisting distortion of the symmetry might very well be supposed to occur.

for on the hypothesis proposed by Barlow and Pope it is difficult to see. On the other hand, the flexibility of interpretation allowed by the theory of compressible atoms is amply sufficient to account for just this sort of thing. Similar small but significant differences in isomorphous crystal shapes are pointed out in a recent interesting paper by B. Gossner (Krystallographie, xliv., 417). Here again the evidence that in isomorphous mixed crystals the two components mutually influence each other in such a way that their properties are no longer those of the pure state is an outcome which might have been definitely predicted from the point of view of the theory of compressible atoms.

The data presented by Groth's "Chemische Krystallographie" (Engelmann, Leipzig) concerning not only inorganic but also a great variety of organic substances will enable anyone interested to test for himself these statements. In general, it will be seen, as has been pointed out above, that the rigid application of the theory of definite valency-volume is impossible and that according to it many facts such as those given in Repossi's table above are inexplicable.

The cases cited by F. M. Jaeger (Journ. Chem. Soc., 1908, xciii., 517) in favour of the Barlow-Pope theory may be used (in so far as they show significant parallelism) equally_well to support the theory of compressible atoms. The two theories are similar, as had already been pointed out, in the ideas of close-packing and definite marshalling. Hence in these and in other cases the arguments with regard to these two matters apply about equally to each; but when Jaeger calls the crystals of phthalimide and saccharin closely similar in form, one cannot help thinking that the argument is somewhat farfetched. Here follow the actual crystallog1aphic data :

The axial ratios in the camphor series presented by Barlow and Pope a few pages afterward show somewhat the same sort of treatment. The first point that strikes one upon comparing the figures is that hexagonal cam. phor and its orthorhombic substitution products are very different from one another in their actual crystal forms and that the dibromcamphors differ much among themselves. Barlow and Pope have been interested in devising a mathematical method which should reconcile these figures; but if the method had been less arbitrary, it❘ Saccharin would have been more convincing. In order, for instance, that the orthorhombic dibromcamphors may be reconciled at all with hexagonal camphor, arbitrary multiples of the axes must be chosen, one of them being as complicated as 5/3. One cannot but wonder if any similarity in crystal. line form would have been detected between these substances if one had not happened to know beforehand that they had a similar structure. Grave doubts may well arise as to the legitimacy of turning away thus from the indication of the actual external crystalline forms.

It will be clear from the foregoing statements that the interest of these investigators has been primarily to show analogies of crystal form. While this is very suggestive, to me the differences seem to be even more interesting than the analogies; for the differences give us a clue whereby further insight into the internal nature of crystals may be obtained.

There are many other cases of approximate similarity of form, where it is hard to see how Barlow and Pope's theory may be made to apply; for example, the interesting series of nitro-dihalogen benzenes recently studied by E. Repossi (Zeit. Kryst. Min., xlvi., 202; through Chem. Zentralb., 1909, xi., 273). A table giving the axial ratios of four of these compounds follows:

1-Nitro-2.5-dibromo benzene

13854 I: 0.7876

I-Nitro-2-chloro-5-bromobenzene 1.3823: I: 0.8196 I-Nitro-2-bromo-5-chlorobenzene 14159: I: 0.8157 1-Nitro-2.5-dichloro-benzene 1'4385 I 0·8223 It will be noted that there is a very distinct difference in the fundamental axial ratios of these four highly analogous substances, amounting to several per cent. Clearly the introduction of chlorine causes a relative shortening of the

axis b even more evident than that shown in the substituted

* Fournal of the American Chemical Society, xxxv., No. 4.

Phthalimide

These marked differences in axial ratios, amounting to 10 per cent in the case of a, must correspond with real differences in internal conditions; the extra oxygen in saccharine and the substitution of sulphur for carbon are by no means unimportant matters. The method of "equivalence parameters" reduces these data to much closer agreement; thus x, corresponding to the axis of a, is found to be 2.518 in one case and 2:462 in another. The difference is only about 2 per cent instead of 10. Such a circumstance cannot convince me that the crystals are to be considered as alike; I am more inclined to believe that the outcome gives evidence of the efficiency of the mathematical method in hiding real differences in crystal forms.

Jaeger's earlier statement that the Barlow-Pope theory is not to be regarded as affording an explanation of valency, nor as a substitute for the theory of structure of organic chemistry, nor as furnishing an indication of the occurrence or non-occurrence of miscibility among solid phases, seems to have more justification. The single advantage cited by Jaeger, that the Barlow-Pope hypothesis enables one to predict similarities in crystalline form, would apply equally well to any reasonable theory inif some elasticity in these spheres is allowed-and the volving definite close-packing of the spheres of influence, facts seem to demand the admission of the existence of this latter property.

Many other cases are considered by Barlow and Pope, some of them with great ingenuity, and all with profound

knowledge of crystallography. Some of the arrangements, are very plausible, but most of them would hold quite as well if not better according to several other hypotheses as to the relative volumes of the several atoms, especially if we take into account atomic compressibility. The explanation for the similarity in crystalline form of sodium nitrate and calcium carbonate is especially interesting and ingenious, because, according to their theory, the former has a different total valency-volume from the latter. This similarity also can be explained in other ways than on their basis. The well-known older explanation, supposing the similarity of form to be due to the fact that the two molal volumes are almost identical and the number of atoms in each the same, is quite as convincing as Barlow and Pope's more elaborate theory (Fourn. Chem. Soc., 1908, xciii., 1528). Indeed this older explanation fits in very well with the idea defended by the present paper; for the attribute of atomic compressibility is all that is needed to explain the indisputable fact that two somewhat differently combined but otherwise analo gous groups of atoms of approximately but not exactly equal bulks can be pressed into very similar forms.

There are other cases of isomorphism which seem to be wholly beyond the reach of the valency-volume theory of Barlow and Pope-for instance, the familiar isomorphism of ammonium and potassium salts. Here the idea of valency-volume gives ammonium nine volumes, but potassium only one; and it is hard to see how any sort of analogous symmetry could be constructed in the two cases under these circumstances. On the other hand the theory of compressible atoms plausibly suggests that the five atoms making up the radical ammonium, which seem together to possess about the same volume as potassium (the molal volumes of potassium and ammonium chlorides are respectively 37.8 and 35′0 cc.), must be compressed by their mutual affinities until a shape not unlike that occupied by the compressed and distorted potassium is

obtained.

It is very questionable if at present there is any use in attempting to make models of the molecules of very complex substances, in view of the many variables which seem to exist within their structure. Because of the fact that sometimes exceedingly complicated structures, such as crystallised ammonium alum, assume very simple crystallographic forms, one is inclined to proceed cautiously with them in his attempt to represent pictorially the actual arrangement of the atoms in a crystalline structure.

Nevertheless, it seems clear, at least to the writer, that the tenets of the theory of compressible atoms give the most reasonable point of attack in a research of this kind. For the theory not only postulates causes which do not seem to be at variance with common sense, but it also gives just that sort of flexibility in atomic volumes and in the mutual adjustment of the effects of the several affinities which seems to be indicated by the slightly differing angles and the varying relations of isomorphous but not exactly identical crystalline forms.

Summary.

In this brief paper the following points are emphasised:

1. The assumption of valency-volume made by Barlow and Pope in 1906, in spite of its great ingenuity, is shown to be not the most reasonable explanation for the actual molecular volumes of solids.

2. The cases presented by them as arguments may be explained quite as well (or often better) in other ways, and therefore afford no arguments in favour of their assumption as against others.

3. Some facts seem to be quite beyond the reach of their hypothesis.

4. In 1902 the theory of compressible atoms indicated a more reasonable method of explaining the tridimensional relations of material, including the symmetry of crystals.

5. This theory assumes, like Barlow and Pope's later

hypothesis, the close packing of the atoms in solids, whereby the form is maintained and the rigidity caused. It proposes, however, that the atoms are compressible and that their volumes are not arbitrary, but depend upon the pressure to which they are subjected.

6. The theory of compressible atoms explains the usual forms of elements and binary compounds better than Barlow and Pope's.

7. The theory shows why elements forming isomorphous compounds need not have exactly the same atomic volume.

8. The theory affords a conceivable picture as to how potassium and ammonium can replace one another isomorphously, a problem apparently beyond the reach of the hypothesis of constant valency-volume.

9. In general, none of the facts cited by Barlow and Pope, and no other crystallographic facts known to the author, conflict with the postulate of atomic compressi bility.

10. Complex compounds possess too many variables for satisfactory treatment at present; but in so far as interpretation is possible, the theory of compressible atoms seems to apply:

SYNTHETIC RESINS.*

By L. V. REDMAN, A. J. WEITH, and F P. BROCK.

Condensation Products of Phenolic Bodies with
Hexamethylene Tetramine.

THE condensation products or synthetic resins which
occur when phenolic bodies are heated in a water solution
of formaldehyde, their polymers or equivalents, are already
well known from scientific and patent literature. The
term "phenolic bodies" signifies any substance containing
a benzene nucleus and having a hydroxyl attached to the
ring, e.g., the cresols, naphthols, thymol, carvacrol,
&c., the chlor- and brom-phenols, nitro-phenols, phenol
sulphonic acid, &c. The term formaldehyde includes its
hydrates or polymers and may be replaced in the reaction
by acetaldehyde benzaldehyde, &c., in certain cases, but
the formaldehyde is in every case more reactive than the
substituted aldehyde or the aldehydes higher in the series.

The formaldehyde-phenol reaction products are formed in every case with the elimination of water as a by-product. The first step in the reaction goes according to the following equation:- C6H5OH + CH2O → HŎCH2.C6H4OH, forming oxybenzyl alcohol. The second change may be represented by the equation:— 2HOCH2.C6H4OH →

HOCH2.C6H4.OCH2.C6H4OH+H2O, in which two molecules of oxybenzyl alcohol unite with the elimination of water and the formation of saligenosaligenin. Further reaction may occur in which the saligenin molecules unite, forming a saliretin product with the further elimination of water.

The water of reaction is not the only water present, since commercial formaldehyde is 60 per cent water, and in the process of formation this water separates out from the newly formed resin.

The wet process for obtaining the synthetic resins has been exploited successfully, commercially, by a number of research chemists who have patented the results of their researches.

Difficulty is experienced in following the rate of this wet reaction, for, heretofore, phenol has been difficult to determine in the presence of formaldehyde, and formaldehyde requires a considerable length of time (about fortyeight hours) for each determination, the results being unreliable within 2 or 3 per cent. Consequently, great

* Journal of Industrial and Engineering Chemistry, vi., No. I.