BEAM, 6 ins. CAPACITY, 200 grms. SENSITIVENESS, 0.5 mgrm. Agate knife-edges and bearings.

Price

£7:15:0.

Illustrated Catalogue Po t Free. TURNMILL ST.. LONDON.

Required for abroad, a Gentleman specialised

in the manufacturing of Chloride of Lime, for erecting and setting in working order of a factory for the making of Chloride of Lime ("Weld n")-Send offers under No. 61,603 to the Agence Centrale d'Annonces, L. and E. METZL and CIE., St. Petersbourg, Morskaia 11 (Russia).

Wan

anted.-Bomb Calorimeter for Solid and Liquid Fuels, Assay Balance, Assay Furnace, Viscosimeter, Thermometers 1/100, Miscellaneous Chemical Apparatus, Glass, &c., Platinum Apparatus.-Address, B. C., CHEMICAL NEWS Office, 16, Newcastle Street, Farringdon Street, London, E.C.

Metallurgical Chemist of Sheffield, with five

years' experience in large Steel Works, desires to manage small laboratory. Accustomed to all kinds of Steel Works Analysis. Age 22. Moderate salary. Completed University course.-Address, W. S., CHEMICAL NEWS Office, 16, Newcastle Street, Farringdon Street, London, E.C.

SULPHUROUS

ACID and SULPHITES.

Liquid SO, in Syphons, for Lectures, &c.

PHOSPHORIC

ACID and PHOSPHATES.

For Chemical and Bacteriological Research.

for all purposes.

4. BOAKE, ROBERTS, & CO. (LIMITED),

Stratford, London,

with hydrogen peroxide (Ber., xli., 1514; D.R.P., 224953), but the arsenoxides can be oxidised by iodine (Ber., xliii., 914) in acetic acid solution, and the arsendichlorides by chlorine :

RASC12 RA8C14 → RAsO(OH)2.

The oxidation of the arsenoxides by iodine is quantitative, and can be used for the estimation of these compounds. Properties.-The arsonic acids are well crystallised compounds which, as a rule, are difficultly soluble in cold water but easily soluble in hot water. They form easily soluble salts with the alkali metals, and these salts are neutral in reaction and have the formula ArAsO (OH) (ONa). The salts are best isolated by adding alcohol to their aqueous solutions. The salts with the heavy metals are insoluble, and in them, as a rule, both hydrogen atoms are replaced, e.g., ArAsO(OAg)2. Arsanilic acid gives a mercury salt (D.R.2. 237787) of the formula

This salt is insoluble in water, but is soluble in salt solution and in glycerin.

The magnesium salts are precipitated when the arsonic acids are boiled with magnesia mixture, but are not formed in the cold (Fourn. Am. Chem. Soc., xxxiii., 101). This behaviour forms a convenient test for the presence of free arsenic acid in arsonic acids.

The arsonic acids when heated frequently lose a molecule of water and pass into anhydrides (Annalen, cci., 205) of the formula ArASO2. These anhydrides, or arsino-compounds, do not absorb moisture from the air, but pass back into the arsonic acid when treated with

warm water.

The arsonic group is, as a rule, firmly attached to the molecule, but can be replaced by halogen by treating with halogen acid. This reaction forms the standard method of determining the position of the arsonic group.

The arsonic acids are stable to oxidising agents and on reduction pass into arsenoxides, arseno-compounds, or primary arsines, according to the reducing agent used.

The esters (Annalen, cccxx., 194) have been prepared by the action of alkyl iodides on the silver salts. As a rule they are liquids, which are at once hydrolysed by water: ArASO(OR)2,2H2O = ArAsO(OH)2,2ROH.

The aminoarylarsonic acids form the starting out point for the preparation of a very large number of derivatives. Thus the amino-group can be diazotised (Ber., xli., 931, 3859 ; Fourn. Chem. Soc., xciii., 1893; D.R.P., 205775, 215251), and then replaced by a variety of other groups such as halogen (Ber., xli., 1856), hydroxyl (loc. cit., D.R.P., 223796), hydrogen, &c. (Ber., xiiv., 3305), in the usual way. Thus arsanilic acid can be made to yield p-nitriloarsonic acid, and this on hydrolysis gives the corresponding carboxy-arsonic acid. The amino-group can also be replaced by the arsonic group by means of Bart's reaction, thus rendering it possible to obtain poly-arsonic acids. The amino group can be acylated (Ber., xl., 3296, xliv., 3092; D.R.P., 191548, 231969) in the usual way, and when thus protected side chains can be oxidised to car boxyl by means of alkaline permanganate (Ber., xli., 931, 3859). The protected amino-compound can also be nitrated. In this case experience has shown that whereas the acetyl derivatatives are almost useless, the oxalyl derivatives and the urethanes usually give excellent results (Ber., xliv., 3092; D.R.P., 231969, 232879). By this means arsanilic acid can be converted into 3-nitro 4-amino-phenyl arsonic acid, and this on reduction with sodium amalgam in methyl alcoholic solution (Ber., xli., 1656; D.R.P., 206344), or with neutral sodium hydrosulphite (Ber., xliv., 3092), or with ferrous chloride (Ber., xliv., 3300), or with ammonium sulphide (Ber., xli., 1656; D.R.P., 206344), gives the corresponding o-phenylene diamine arsonic acid. This gives quinoxalines when reated with diketones such as phenanthraquinone.

If the nitro-arsanilic acid is diazotised and the diazo compound warmed to 18° with excess of sodium acetate or other substance capable of combining with hydrochloric acid, the nitro-group is replaced by hydroxyl (D.R. P. 244166; Ber., xliv., 3578); a reaction very similar to that which takes place when the diazo-salts from the corresponding nitro-sulphanilic acid are similarly treated (D.R.P. 138258). If the diazo-compound thus formed is coupled with a phenol, such as B-naphthol, and the azo dye thus formed reduced, 3-oxy-4-amino-phenyl arsonic acid is obtained :

More vigorous reduction of the azo-compound leads to the arseno-compound ("Salvarsan”).

The presence of the nitro-group in the molecule loosens the arsonic group. Thus if the diazo-salts of nitroarsanilic acid are boiled, the arsonic group is replaced by hydroxyl (Ber., xliv., 3450, 3578). This loosening of the arsonic-group is remarkable when it is remembered that The the nitro-group is present in the meta-position to it. amino-group is also less firmly bound than usual as it is readily replaced by hydroxyl by the action of aqueous caustic potash (D.R.P., 235141; Ber., xliv., 3449), at 80°.

The amino-group in arsanilic acid, &c. can be methylated in the usual way (Ber., xli., 1514). By treatment with chloro-formic ester urethanes are obtained (D.R P. 232879), whilst chloracetic acid leads to glycines (Ibid., 204664). These glycines can also be obtained by treating the amino-arsonic acid with potassium cyanide and formaldehyde and then hydrolysing the nitrile :

Compounds of a similar nature can, of course, be obtained by using other aldehydes, the reaction being the ordinary Bucherer and Knoevenagel process.

Unsymmetrical ureas and thioureas can be obtained (D.R.P. 213155) from the amino-arsonic acids by means of cyanates and thiocyanates, and thiol-arsonic acids have been prepared by treating the diazo-salts with xanthates and then hydrolysing the xanthic ester formed (D.R.P., 216270). Azo-dyes derived from the arsonic acids have also been described (D.R.P., 212018, 212304, 216223, 222063; Journ. Chem. Soc., 1893. xciii.).

The phenolic arsonic acids can be obtained, as stated above, by fusing the phenols with arsenic acid, but the yields by this method are usually very poor. They are best obtained from the amino-compounds by boiling the diazo salts with water (D.R.P., 205775, 215251, 223796; Ber., xli., 931, 1854 ; lxiv., 3305, 3863). The hydroxyl group loosens the arsonic group. Thus if attempts are made to couple p-oxy-phenyl arsonic acid with diazo

CHEMICAL NEWS, Į Feb. 7, 1913

63

The tetra-chloride thus formed on heating with excess of the arseno-compound gave the dichloride. This on hydrolysis and loss of water gave the arsen-oxide :

Formation. These appear as side products in the preparation of arsonic acids by heating primary bases with arsenic acid (Ber., xli., 2367; Journ. Chem. Soc., xciii., 1180). They are more readily obtained by heating the arsonic acid with excess of primary amine, and by this means arsinic acids containing two different aryl groups can be obtained:

NH2C6H4A5O3H2+ AINH2 =

= (NH2C6H4) (NH2Ar)AsOOH+H2O. The yields, however, are always very poor, and as a rule do not exceed 3 per cent. The acids are more readily obtained by Bart's reaction (D.R.P., 250264; Pat. Anm. Kl. 12 0. B., 69659). This consists in treating a normal, or better, an iso-diazo salt with an arsenoxide in neutral or faintly alkaline solution, with or without the addition of copper powder or copper salts:

ArN:N.Ac +Ar'As(ONa)2 = ArAr'ASONa + NaAc+N2. Properties.-The arsinic acids form salts of the structure R2ASOOM, where M is a monovalent metal, and also exhibit feeble basic properties. Thus diphenyl arsinic acid forms a nitrate (Annalen, cccxxi., 151), Ph2As(ONO3)O, when treated with a mixture of concentrated nitric and sulphuric acids, and dibenzyl arsinic acid forms a chloride, (Annalen, ccxxxiii., 89), Bz2As(OH)2Cl.

The arsinic acids are much less important than the arsonic acids, the best known member of the series being cacodylic acid, Me2AsOOH. They are less soluble than the corresponding arsonic acids, give no anhydrides, and on reduction pass into the diaryl arsenoxides, (R2As)2O, of which very little is known at present.

The Arsenoxides.

Formation. The arsenoxides can be obtained by the action of sodium carbonate on the arsendichlorides. This method is only of importance where the arsendichloride can be readily obtained. For example, it is the easiest method of obtaining p-dimethyl-amino-phenyl arsenoxide, as the corresponding arsendichloride is very readily formed by the action of arsenic trichloride on dimethyl aniline (Annalen, cclxx., 139; Ber., xli., 1514). As a rule, however, the arsenoxides are best obtained by the moderated reduction of the arsonic acids (Ber., xliii., 917; D.R.P., 206057, 212205, 213594, 216270, 235391, 235430). The reduction can be carried out by sulphurous acid, hydriodic acid (in which case sulphur dioxide should also be used to reduce the iodine liberated), phenyl hydrazine, thionyl chloride, or phosphorus tri-chloride. Of these hydriodic acid in the presence of sulphur dioxide gives the best results. The arsen-oxides can also be obtained by the oxidation of the arseno-compounds by means of chlorine or hydrogen dioxide. Thus Michaelis and Loesner (Ber., xxvii., 267) obtained m-nitro-p-amino-phenyl arsenoxide from the corresponding arsonic acid by the following indirect method. The arsonic acid was first reduced to the corresponding arseno-compound by means of neutral

AsO

Properties.-The arsenoxides can usually be obtained crystalline, although as a class they show considerable tendency to form waxes. They are insoluble or only very slightly soluble in water, and melt at a lower temperature than the corresponding arsonic acids. They show but tive groups are present in the ring, such as nitro- or slight tendency to unite with water unless strongly negacarboxy-groups. They dissolve easily, however, in caustic alkalis, but only with difficulty in sodium carbonate (distinction from arsonic acids) to form salts RAS(ONa). Esters of the type ArAs (OR)2 have been isolated by Michaelis by treating the arsendichlorides with sodium alkylates (Annalen, cccxx., 826; cccxxi., 143). They are extremely easily hydrolysed by water.

The arsenic in the arsenoxides is less firmly attached to the molecule than is the case with the arsonic acids. Thus p-aminophenylarsenoxide on prolonged standing or boiling with hydrochloric acid is decomposed (Ber., xliii., 923) into aniline and tri-amino-triphenyl arsine.

Thus when the arsenoxides are treated with sulphuretted The oxygen of the arsenoxide group is fairly reactive. hydrogen, sulphides are formed, e.g., ArAs=S, and treatment with concentrated halogen acids leads to the formation of arsen dihalides. No condensation products with hydroxylamine or hydrazine have been described, but Michaelis has prepared phenyl arsenimide, PhAs=NH, by treating the dichloride with ammonia (Annalen,

CCCXX., 291).

As would be expected from compounds containing trivalent arsenic, the arsenoxides unite with halogen to form oxy-halides, ArAsOCl2, and these on hydrolysis pass into the arsonic acids. The arsenoxides can also be oxidised to the arsonic acids by hydrogen dioxide, and by iodine in acetic acid solution (Ber., xliii., 917). This latter reaction is quantitative, and can be used for titrating arsenoxides.

The arsenoxides are sufficiently powerful reducing agents to reduce Fehling's solution when boiled. On reduction with metal and acid, stannous chloride, sodium amalgam, phosphorous acid or sodium hydrosulphite in neutral solution, they pass into the arseno-compounds. (To be continued).

Relative Electric Conductivities and Ionisation of Aqueous Solutions of Hydrochloric Acid.-J. A. Muller. The author has determined the electric resistances of aqueous solutions of hydrochloric acid of various concentrations at 18°, 51°, and 81°, using Kohlrausch's method. From the results thus obtained he has deduced the coefficients of ionisation of solutions of the acid, and has found that ionisation diminishes rapidly as the con centration increases, but for the same concentration it varies only very slightly with the temperature.-Bull. Soc. Chim. de France, xi.-xii., No. 23

LOWERING OF VAPOUR PRESSURE.

(A SIMPLE DEMONSTRATION).

By W. W. REED, M.Sc., F.I.C.

was the greater. He is careful to state, however, that this is no absolute proof that the density of a substance in a powdered condition is greater than that of a block of the same substance, since we cannot be sure of the complete identity of the substance in the two forms. Indeed, we now know that in very many cases the form of a substance when precipitated is different from that observed in nature or obtained by fusion. Rose's work, therefore, affords no information on the influence of the state of division of a substance on its specific gravity.

ON THE DENSITY OF SOLID SUBSTANCES, WITH ESPECIAL REFERENCE TO PERMANENT CHANGES PRODUCED BY HIGH PRESSURES.* By JOHN JOHNSTON and L. H. ADAMS. (Continued from p. 57).

Effect of Powdering a Solid upon its Density. MORE than a century ago, Hassenfratz (Gilbert's Ann., 1799, i., 369; Ann. Chim., No. 27, 188) published a paper in which he claims that by breaking up a piece of glass weighing about 50 grms. into 2520 pieces (how this was accomplished without loss of material is not stated), a change in density ensued; but his experiments are in part discordant among themselves and are by no means accurate enough to decide the point at issue. His work may therefore be left out of account entirely; it has been mentioned here solely because its conclusions have been quoted occasionally.

The second paper, in point of time, dealing with this question, is one by G. Rose (Pogg. Ann., 1848, lxxiii., 1); it deserves mention if for no other reason because it has been widely cited and frequently by authors who apparently have not read through the original. Rose worked, on the one hand, with gold and silver in the massive state, and with natural barite; on the other hand, with powders produced by precipitating these substances by chemical agency; and he found that in each case the density of the powder From the Journal of the American Chemical Society, xxxiv.,

No. 5.

The Earl of Berkeley (Journ. Chem. Soc., 1907, xci., 60) has measured carefully the densities or two sizes of fragments of natural barium sulphate. For particles remaining on a sieve, the openings of which were 0'57 mm. on a

of particles remaining on a sieve with 0-35 mm. openings averages 4'4700. The difference between the two densities is less than 0.005 per cent-within the error of experi ment. The difference between the sizes of the particles in the two cases, however, is so comparatively small that the evidence here adduced is insufficient to decide the point at issue.

It may be noted, in passing, that the above work of Rose has been cited (Cameron and Bell, Bull. 30, 1905, Bureau of Soils, Department of Agriculture, p. 43) in support of an observation of Spring (Mémoires Soc. Geol. Belg., 1903, xvii., 13 33), according to which the amount of water required to fill the spaces of a given mass of sand is greater than would be expected if it were merely a phenomenon of occupying the air spaces. Spring's experimental work on this point is altogether insufficient to decide the question; his conclusion is based on the assumption that 26 per cent of the total volume occupied by the sand consists of air spaces; even then the difference observed was only o·8 cc., or but 0.25 per cent of the total volume of the sand. In what follows it will be shown directly that this conclusion is false, since the density of solid substances, determined by the pyknometer, is found to be independent of the size of the particles, so long as these are strictly homogeneous, or even to show a slight decrease as the size of particle decreases; whereas, according to Spring's conclusion, the density should increase as the size of the particles decreases. A number of other early papers are cited by Spring in his first paper (Bull. Acad. Roy. Belg., 1883, 131, vi., 507) dealing with the change of densities of solids; these are all, however, concerned with changes of density produced by crystallisation, vitrification, by tempering, hammering, or annealing, and will be considered later.

Our experiments were made with quartz and potassium sulphate. The material was passed through a series of sieves, rated as 40, 60, 80, 100, 120, 150, and 200 meshes to the linear inch; the various samples were collected and dried at 200° for half an hour (special experiments showed that the density of even the finest material was not appreciably affected by exposure to the air for 48 hours). The earlier determinations of densities were made with the old style of pyknometer; the later ones with the new style described above. The liquids used were water and xylene (for which d4/30=0.85262), the latter being used exclusively

with the salts.