Free potassium cyanide Potassium carbonate

...

26 14 grms. per litre. 47.25 165.00 99

A photomicrograph of a deposit from this bath made with a current density as high as 8 amperes per square foot showed the same structure as a deposition obtained under commercial conditions with a current density of 3 amperes per square foot.

The reduction of time in deposition, resulting | in double the output, should be of considerable value, and the enormous saving in capital outlay in materials and plant will be appreciated by electroplate manufacturers.

"The Crystalline Structure of Electro deposited Silver. By G. B. BROOK, F.I.C.

The author shows that the normal relatively smooth "reguline" deposit develops, with high current density, a growth of individual acicular crystals from the face of the deposit. In the subsequent burnishing and polishing such structure results in the retention in the interstices of the imperfectly laid-down crystals of the polishing material. It was the investigation of this latter trouble, which had become serious and general in the silver trade, that the author found that very high current density contributed to the particular form of crystal growth (or its free development) and accounted for the retention of the rouge in the final process of polishing.

"The Deposition of Gold-Silver Alloys." By S. FIELD, A.R.C.Sc.

In

A series of experiments has been carried out in order to trace the influence of varying conditions on the composition of the gold-silver alloys deposited in the well-known "green gold." In cyanide solutions gold is the more positive metal. In the absence of hydrogen deposition, the composition of the deposited alloy can be calculated from the ratio of its mass to that of copper deposited in a coulomb-meter in the circuit. gilding solutions of normal strengths this ideally quantitative deposition is not attained. The deposits are produced on lead foil cathodes, and, after weighing, are cupelled with added silver to give a "parting" mixture from which the content of gold is determined. Platinum anodes provide the means of most constant control over the composition of the bath, additions of standard solutions of gold and silver cyanides being made after each deposit.

A cold solution containing 142 grms, gold and 7.2 grms, silver per litre and without free cyanide was first employed. Increased current density increased the proportion of gold in the alloy. A similar increase occurs in a warm solution but the whole of the gold values are lower than in cold solutions. Diluting the solution also increases the proportion of gold and addition of free cyanide has the same effect, the proportion of gold to silver becoming more constant and independent of other conditions with larger proportions of free cyanide.

By way of comparison, deposits were prepared in a works under normal working conditions. These included a more dilute solution with a

larger gold-silver ratio and a larger proportion of free cyanide. Over a range of current density of one to five an almost constant percentage of 77 to 78 per cent of gold in the deposit was obtained.

NOTES.

LINSEED OIL SUBSTITUTES.-Different combinations of Guiana wood oil, fish oil, and soja oil have been utilised with variable success (W. Bucchlans, Seifenseider Zeitung, 1920, 2, xlvii., 141). A method of manufacture consists in heating a fine quality fish oil for 2 hours at 240° C., then distilling with steam at 375° to 400° C. during 30 hours. Thus, a thick elastic material is obtained which has very good siccative properties. Another method is to heat the oil with 15 per cent water and o5 per cent sulphate of manganese in an autoclave during 3 to 5 hours at a pressure of 10 atmospheres. The conclusion arrived at, according to Bruckhams (Seifenseider Zeitung, 1920, xlvii., 141) is that practically it is impossible to find a substitute for linseed oil, especially in manufacture of lakes. The mixed varnishes are made with boiled soja, sunflower, and linseed oils. If one wishes to give siccative properties to non-siccatives such as fish oils, it suffices to mix a drop of glacial acetic acid with them in the tared tube of a centrifugal machine; then all is cooled at -10° to o° C., and 1 to 12 cubic cm. of bromine added drop by drop. The mixture is left in the refrigerant for 2 hours, and then worked in the centrifugal apparatus. The ether is decanted and the residue recovered in the tube with 2 cubic cm. of cooled ether and again passed through the centrifugal apparatus. The operation is repeated, the precipitate dried at 110° C., and weighed. The percentages of hexabromides obtained with different linseed oils may vary to 8 per cent. This is why Bruckhams prefers this method to that of Steel and Washburn, which gives 3 per cent smaller results, but usually constant even for different oils.-Rev. Chim. Indusrtr., September, 1920.

on

ROYAL INSTITUTION.-The following are the Lecture arrangements at the Royal Institution, before Easter: Prof. J. Arthur Thomson, a course of lectures on the "Haunts of Life," adapted to a juvenile auditory, to begin on December 30; Sir Gerald P. Lenox-Conyngham, two lectures on the "Progress of Geodesy in India"; Sir James G. Frazer, three lectures on "Roman Life (Time of Pliny the Younger)", "London Life (Time of Addison)," "Rural English Life (Time of Cowper)"; Dr. Arthur Keith, four lectures "Darwin's Theory of Man's Origin"; Dr. W. A. Herdman, three lectures on "Oceanography"; Mr. Frank Balfour Browne, two lectures on "Mason Bees and Wasps"; Dr. George C. Simpson, two lectures on the "Meteorology of the Antarctic"; Dr. Percy C. Buck, three lectures on the "Madrigal," with musical illustrations by the English Musical Singers; Prof. A. Fowler, three lectures on "Spectroscopy"; and Sir Ernest Rutherford, three lectures on "Electricity and Matter." The Friday evening meetings will commence on Jan. 21, when Sir James Dewar will deliver a discourse

on "Cloudland Studies." Succeeding discourses will probably be given by Sir Frank Benson, Dr. A. D. Waller, Dr. F. W. Aston, Mr. Solomon J. Solomon, Dr. John Buchan, Sir Frederick Bridge, and other gentlemen.

CHRISTMAS LECTURES AT ROYAL INSTITUTION. Prof. J. Arthur Thomson will deliver the first of his series of Lectures on "The Haunts of Life" at the Royal Institution on Thursday, December 30, at 3 p.m.

New Patents.

THIS List is specially compiled for Chemical News by Messrs. Rayner & Co, of 5, Chancery Lane, London, from whom all information relating to Patents, Trade Marks and Designs can be obtained gratuitously.

Mercuric Oxide.-Mr. C. Brusa and Messrs. Borelli & Co., of Turin, Italy, have recently obtained a patent in this country for means for obtaining mercuric oxide No. 15091. The apparatus for which, is illustrated in the appended drawings. Mercuric oxide is obtained by decomposing mercurious nitrate by heat in a closed chamber; the nitrogen oxides evolved are brought into contact with mercury in presence of oxygen to yield more mercurious nitrate for use in the process. The apparatus shown consists of two revolving bull-mills, A, B, connected by a hollow trunnion, c; a charge of mercurious nitrate is placed in the mill A, and mercury in the mill B, which latter is provided with an oxygen inlet-pipe, G; the mill A is heated and the mill B cooled. so that nitrogen oxides are evolved from the former, leaving mercuric oxide, and pass through the shaft c into B, where they react with mercury and oxygen to reform mercurious nitrate.

NOTICES.

EDITORIAL.-All Literary communications and Books, Chemical Apparatus, &c., for review or notice to be addressed to the EDITOR.

SUBSCRIPTIONS, £1 12s. per annum, payable in advance, should be addressed to the MANAGER.

BACK NUMBERS and VOLUMES can be purchased on application to he MANAGER.

THE CHEMICAL NEWS,

97. SHOE LANE, LONDON, E.C 4.

ADVERTISEMENTS.

All communications for this Department should be addressed toT. G. SCOTT & SON,

63, LUDGATE HILL, LONDON, E C.4.

DRYING APPARATUS. The owner of

British Patent 114620 quick drying apparatus principally for

laboratory work, is willing to sell the Patent rights or to grant licences to manufacture on reasonable terms. Apply for particulars to Feeny & Feeny, Chartered Patent Agents, 60, Queen Victoria Street, London, E.Č.4.

AN old-established small practice in London

for sale, suitable well trained young Analyst, or would consider junior partner.-Apply Box 820, c/o Scott & Sos. 63, Ludgate Hill, E.C.4.

PATENTS, TRADE MARKS. Handbook

and advice free-B. T. KING, British and U. S. Regd. Patent Attorney, 146a, Queen Victoria St., London. 35 years reference

If in good condition, Sixpence per copy will be

paid for any of the undermentioned numbers of the CHEMICAL NEWS which may be forwarded to this office:

3048, May 24th, 1918. 3051, July 5th, 1918. 3053, August 2nd, 1918 3054, August 16th, 1918.

3056, September 13th, 1918.

3059, October 25th, 1918 3062, December 6th, 1918.

3063, December 20th, 1918.

3064, January 3rd, 1919.

3066, January 17th, 1919.

and hydrogen. Dec. 7th.

Volumes of the

THE principal applications of nitrocellulose in industry are to celluloid, propellants, artificial silk, colloidion, and some blasting explosives. In this paper I shall confine myself for the most part to nitrocellulose as used in the manufacture of propellants, and I propose to describe some of the relations that exist between the solutions of cellulose and of nitrocellulose and the gelatinised nitrocellulose after the solvent has been eliminated, together with some work on the characteristics of its solutions.

Most of the modern propellants are colloidal in | character, and it is to the nitrocellulose which they contain' that this character is to be attributed. As they have no definite crystalline structure, and can be obtained as uniform non-porous masses by the process of incorporation or malaxation during manufacture, they present an unbroken surface to the flame which ignites them, and even under pressure continue to burn regularly from layer to layer. It is this property that has enabled guns to be designed much lighter in weight than for gunpowder, and that has made it possible by suitable choice of thickness of the propellant to secure very high velocities of projectiles (within a narrow margin of a few feet per second from round to round) without undue strain on the gun.

To obtain the degree of regularity in ballistics necessary for accurate shooting it is important for the dimensions of the finished propellant to be within very narrow limits. It has been found that the nature of the nitrocellulose used affects the amount of shrinkage which takes place in the cord after squirting from the die, and it is obvious therefore that the colloidal properties of nitrocellulose gelatinised by means of solvents, either volatile or non-volatile, are of great importance in the study of propellants.

I propose to bring to your attention the results of some work on gelatinised nitrocellulose that has been carried out at the Royal Gunpowder Factory, and the Research Department, Woolwich.

Among the more important aspects that have to be considered are the influence on the product of the nature of the raw material, the influence of the process which it undergoes in being nitrated and stabilised by boiling with water, and the effect of these on the characteristics of the finished material. Among other important properties the viscosity of solutions of cellulose and of nitrocellulose in relation to these aspects clearly required investigation.

*A General Discussion before the Faraday Society and the Physical Society of London, October 25th, 1920.

Treatments which affect the Viscosity of Cellulose and Nitrocellulose Solutions.

In general, those treatments which would be expected to affect disruption of molecular aggregates (e.g., by grinding or chemical decomposition such as hydrolysis) are found to lower the viscosity both of cellulose and nitrocellulose solutions.

The treatment of the cellulose before solution, with bleaching powder (oxidation) or caustic soda (hydrolysis) in the usual processes of cleaning, reduces the viscosity of its solution in cuprammonium. The viscosity of the solution itself is reduced by exposure to air or light, or even to some extent by simply allowing it to stand for some time. The viscosity of the solution of the corresponding nitrocellulose is found to be affected in the same direction.

With the nitrocellulose, the steam boiling treatments given in the purification process lower its viscosity, but there appears to be a practical limit to the process. When the more easily attacked forms of nitrocellulose have all been destroyed by boiling, the further reduction in viscosity is very slow, and if the cellulose before nitration has been treated so as to reduce its viscosity markedly, no great reduction of viscosity is produced by boiling the nitrocellulose.

In order to produce a nitrocellulose of low viscosity, it is found to be better to reduce the viscosity of the cellulose as far as possible by suitable treatments than to rely on the reduction during boiling of the nitrated product.

It may be mentioned parenthetically on the question of the measurement of the state of molecular aggregation of nitrocellulose in solution, that in a recent paper by Duclaux and Wollman (Bull. Soc. Chim., xxvii., 420), emanating from the Pasteur Institute, while the lower limits of molecular weight of nitrocellulose solutions are shown by measurements of osmotic pressure to be very high, the molecular weight of products of different degrees of viscosity varied from aggregates of 80 times the C, formula for cellulose to 270 times in the case of the more viscous samples. Relation of Viscosity of Solution and of its Nature to the Toughness of the Dried Nitrocellulose Colloid.

In some earlier work the viscosity was measured by means of a viscosimeter of a simple type, and the toughness of the resulting gelatinised dried nitrocellulose in the form of cord was evaluated by determining the least radius of curvature to brought out such differences as that a cord of which the cord could be bent. This method nitrocellulose of 13 per cent nitrogen content cracked at a radius of curvature of 100 millimetres when acetone had been the solvent, while a cord of nitrocellulose of 12:4 per cent of nitrogen broke at a radius of 25 mm. again with acetone as the solvent, and at 5 mm. when ether-alcohol had been used to gelatinise it.

From the results of these experiments it was deduced that when the same simple solvent (acetone) was used to gelatinise nitrocelluloses of a high nitrogen content (13 per cent) the viscosity of the nitrocellulose solution was inversely proportional to the toughness of the resulting cord. The coherence of the dried nitrocellulose was thus found to be adversely affected by the higher state of aggregation of the molecules of the nitrocellulose, as indicated by the viscosity of its solution.

This short note on these experiments has been given, as they illustrate a connection between the viscosity of a solution and the physical consistency of the dried nitrocellulose.

The nature of the volatile solvent also affects the brittleness of the dried gelatinised substance. Thus the dried product made by gelatinising nitrocellulose of 124 per cent nitrogen with etheralcohol is much tougher that that gelatinised with acetone, and a similar effect is obtained by adding some alcohol to the solvent acetone. Caution has, however, to be observed in this connection, as retention of the solvent in an imperfectly dried substance may give a quite fictitious semblance of toughness to a film of gelatinised nitrocellulose. Thus when films of nitrocellulose of 13 per cent nitrogen are made with acetone and with amyl ace tate respectively, clear acetone films of about 0.5 mm. in thickness (and these can be made by taking precautions to prevent access of water vapour during evaporation of the solvent) seem to be much more brittle than similar films made with amyl acetate. When, however, the latter are dried out so that all the solvent, in this case of considerably higher boiling point than acetone, is eliminated, the film is quite as brittle as that in which acetone is the solvent.

Processes for Determining the Viscosity of Cellulose and of Nitrocellulose Solutions.

It might now be well to refer to some of the methods that have more recently been found most suitable for determining the viscosity of solutions of cellulose and nitrocellulose, and to state briefly the information that study of this property has afforded. Of great importance in this connection has been the application of the Falling Sphere Viscosimeter (Gibson and Jacobs, J. Chem. Soc., 1920, cxvii., 472) in which the velocity is observed of a steel ball, o 15 cm. in diameter, under standardised conditions, falling through the solution, appropriate corrections being made for deviations from Stokes' law on account of the side-effect of the walls and the end-effect of the bottom of the viscosimeter tube. This instrument has had a wide application in the study of solutions of nitrocellulose, and the difficulties inherent in its application to solutions of cellulose in cuprammonium solution, chiefly on account of the need for eliminating air in this estimation, have been overcome (Gibson, Spencer, and McCall, J. Chem. Soc., 1920, cxvii., 484).

For cellulose dissolved in cuprammonium solution a capillary viscosimeter in which the solution comes into contact only with hydrogen was developed, and the conditions were obtained for the preparation of the cuprammonium solution by a process which gave more reliable results than those described by Ost. By the use of this capillary viscosimeter it is possible to obtain viscosities of more mobile solutions of cellulose in cuprammonium solution than with the Falling Sphere Viscosimeter, for example, below 10 C.G.S. units. Relation between Viscosity of Nitrocellulose Solution and Cohesion of Dried Nitrocellulose

Mixtures.

Reference has been made to a case in which the brittleness of dried gelatinised nitrocellulose was affected by the nature of the original nitrocellulose and associated with high viscosity of the nitrocellulose solution. A similar relationship

came to light when a new nitrocellulose-nitroglycerine propellant (R.D.B. Cordite) requiring ether-alcohol as a solvent was used in quantity during the war, in place of cordite M.D. for which the supply of acetone was insufficient. It was found on pressing the incorporated gelatinised material that the extruded cord exhibited a lack of cohesion and uniformity of diameter which gave rise to difficulties. An investigation of the nitrocellulose and of the cellulose from which it was made was carried out by the application of the processes mentioned above, and ultimately the cellulose was controlled by the Falling Sphere Viscosimeter. By this means the uniformity of the cellulose was ensured, and a limit imposed on its viscosity. It appeared that the lack of uniformity of the pressed cords of this propellant in this case also was due to an unduly high viscosity of the nitrocellulose, and when measures were taken to control the viscosity of the cellulose itself, no further trouble was experienced.

From a consideration of these two examples, one of gelatinised nitrocellulose and the other of a nitrocellulose-nitroglycerin mixture, it would appear that a high state of aggregation of the molecules of nitrocellulose is not conductive to the toughness that is desirable in a product that has to be handled both during manufacture and after.

The subject of the viscosity of cellulose and of solutions of cellulose was further pursued at the Research Department, Woolwich, by the application of the methods of determining viscosity mentioned above, and some of the results obtained in these investigations may now be considered.

Viscosity of Cellulose Solutions and their
Relation to Nitrocellulose Solutions.

In the first place, as regards the cellulose itself, the reduction in its viscosity brought about by increase in temperature of the alkaline pressure boiling to which the cotton is subjected, and the influence of increase in the concentration of the caustic alkali in this process in the same direction have been described by Gibson (loc. cit., 481).

Correlations were then made between the viscosity of the cellulose in cuprammonium solution and that of the solution in ether-alcohol of the nitrocellulose made from it, and it was established that although the processes of nitration and stabilising by boiling of the nitrocellulose tended to level down differences, yet there is a general relationship in the sense of higher viscosity of the nitrocellulose made from cellulose of high viscosity, and it was brought out that if the cotton had been incompletely reduced in viscosity by the soda boiling the viscosity of the finally purified nitrocellulose was with greater difficulty brought down by the steaming during stabilisation. Minimum Viscosity of Nitrocellulose Solutions. The viscosity of solutions of nitrocellulose in ether-alcohol was then studied (Gibson and McCall, J. Soc. Chem. Ind., 1920, xxxix., 172, T), by the method of the Falling Sphere Viscosimeter, and a suspicion that a small change in the proportion of ether to alcohol affected the viscosity, led to a thorough examination of this variant with ether-alcohol solutions of nitrocellulose of different contents of nitrogen and in different concentrations. It was established that there was a minimum viscosity at a certain proportion of ether to

alcohol, solutions on either side of the minimum being the more viscous, and indeed formed jellies with the higher proportions of either alcohol or ether. The position of this minimum was also found to be independent of concentration, leading to the deduction that this minimum would give the optimum proportions for the solvent used in the process of gelatinisation in the incorporator where the ratio of solvent to substance is necessarily very much less than is any solution of which the viscosity can be measured by the means spoken of above. It may be said that this was verified in practice.

This position of minimum viscosity was found to vary with nitrogen content among those nitrocelluloses which are soluble in ether-alcohol more ether being required for those with the higher content of nitrogen, and it was also established that the farther the composition of the solution is removed from the point of minimum viscosity, the longer does it take for equilibrium to set in, the process of gelation continuing in some of the mixtures rich in one constituent for a considerable time. It proved to be of importance from a manufacturing point of view to choose the mixture of minimum viscosity, since with the associated proportion of ether to alcohol on the one hand, gelatinisation proceeds with the greatest rapidity, and on the other hand, as this mixture has the maximum fluidity, with the smallest usage of solvent. The practical application of this principle in supply, the large scale experimental work carried out in this connection and the means taken to secure that a nitrocellulose of easy gelatinising properties was obtained by the controlled treatment of the cotton itself are described by Mr. R. A. Punter.

The alcohol used in the above work was of 92 per cent strength, and it has been found possible by determining the viscosity of a series of binary mixtures of nitrocellulose in ether and alcohol containing varying proportions of water, to obtain a ternary diagram from which the mixture of ether, alcohol, and water that gives the optimum value of viscosity can be read off.

The

A similar point of minimum viscosity has been found with acetone-water solutions of nitrocelluloses, and it is of interest that the proportion of water in the acetone thus indicated is one which has long been used in the manufacture, having been chosen from other considerations. phenomena of the gelation of such mixtures are discussed by Masson (J. Chem. Soc., 1920, cxvii., 819). The relation between viscosity and concentration, as is also the case with ether-alcohol solutions, appears to conform to the Arrhenius logarithmic expression, but only over a limited range.

Concluding Remarks.

The above work has afforded a great deal of information on the nature of the solution of cellulose and of nitrocellulose and forms the basis for theoretical speculation apart from the bearing it has had on manufacture.

The viscosity of the solutions has been proved to have a direct relation to the properties of the dried colloid from which the solvents have been eliminated by drying. The dried colloid, however, presents difficulties in assessing its physical properties. One method of determining its toughness has been mentioned, and determinations of

hardness by the Brinell method, of breaking strain, or of resistance to crushing have been applied, but not so far with much that is informative. One method may be mentioned by means of which variations in the viscosity of the hard colloid may be traced. If large cords be pressed from different composition of the cordite type in which the proportion of guncotton to nitroglycerine is varied, these cords retain a small proportion of the volatile solvent. If now the quantity of this solvent be determined in different concentric zones it is found that as the nitrocellulose becomes dominant, a proportion is reached when the distribution of the volatile matter is no longer uniform but increases regularly from skin to core, the ratio of the content of volatile matter in the different regions remaining constant although the total quantity is reduced in the process of drying. This resistance to the migration of the solvent is some measure of the viscosity of the horny substance, and its regional determination might, perhaps, assist in the investigation of the physical condition of other hard colloids.

It is hoped that the above will show the importance of the application of the investigation of the colloid solutions of nitrocellulose and indicate, perhaps, points of theoretical interest that remain to be followed up.

ISH COMPOSED OF A WHOLE-NUMBER
PART (4) PLUS AN AUXILIARY PART (B)
AND A ROTATING ELECTRON (C) ?
By F. H. LORING.

READERS of the CHEMICAL NEWS will remember Masson's discussion in these columns (1896, lxxiii., 283) of the position of hydrogen in the Periodic Table. A further discussion will be found in Friend's "Text-book of Inorganic Chemistry," I., 274. Turning now to the Lewis-Langmuir "Octet Theory" (Journ. Am. Chem. Soc., 1919, xli., parplaced respectively over lithium and beryllium ticular p. 874), it will be seen that H and He are in the table shown. If the table is expanded horizontally by removing the following elements (Table I) from their present sandwiched-in positions and arranging them, as here shown, it brings Groups I. and VII, are deficient in members; but into prominence the fact that the lower parts of of course, elements may be discovered eventually which will fill the gaps.

If it be assumed that hydrogen belongs to Group VII. (Masson) and that it also belongs to Group I. (see Friend loc. cit.) the above deficiency is partly made up at the top of the table. It is not necessary for helium to be placed above beryllium, since the former finds its normal place on the extreme left of the table next to Li. argument thus far is not convincing as the absence of elements in one part of the table may have nothing to do with a relative excess at another corresponding part, though shortage of material may apparently be the cause.

The

Now referring to Aston's researches, which have pretty well demonstrated that all the elements except hydrogen have whole-number weights, it is evident that the

atomic electric