into the grand total of 1904 millions of dollars given as the value of the mineral productions of the United States. The same is true of every other metal on the list; in some cases, notably, perhaps, in that of aluminium, the value of the metallic product is many times greater than that of the mineral from which it is produced; thus the value of the aluminium produced is given as 5 million dollars, whilst that of the bauxite from which it is produced is about 450,000 dollars; surely it is the latter figure, and not the former, that should enter into a list of the values of the mineral productions of any country.

In the non-metallic products similar anomalies are also to be met with; cement, bricks, oilstones and millstones are articles that owe a very great, if not in every case the greater, part of their value to the labour, fuel, and power used in their preparation rather than to the crude material from which they are produced; if an American sculptor carves a statue out of native marble, should the value of the finished statue be included in the sum total of the value of the mineral resources? There can only be one answer to such a reductio ad absurdum, and yet the principle is exactly the same as that of including the value of the dressed grindstones instead of that of the sandstone or grit from which they are cut.

The above are matters of principle which present, no doubt, great difficulties in arriving at a satisfactory solution; the coordination of the methods of tabulating the mineral productions of different countries, so as to admit of just comparison, has often been tried, but has never been attained successfully yet, so that all that statisticians can do is to take care that they thoroughly understand the differences that obtain between the various systems in vogue. In other respects the present volumes are quite up to the high standard that we have been accustomed to in the United States Geological Survey publications. As already said, they suffer from want of compression, and there are many repetitions that might be avoided and much superfluous matter that might well be excised. In fact, they require more careful editing than they receive at present, and this is all the more necessary seeing that the different articles are written by different contributors, and are of very unequal value.

For example, no careful editor would pass such statements as we find under the item fluorspar, where we are told that the mineral is "only slightly harder than calcite, and consequently crushes easily," whereas the ease or difficulty of crushing has nothing to do with hardness; and again, "When fluorspar is associated with zincblende, complete separation of the two minerals has been difficult on account of their nearness in specific gravity"; the specific gravity of fluorspar is about 3.1, and that of blende about 4, a difference which should afford an ample margin for successful separation in a suitable appliance.

Finally, it may be pointed out that although these volumes in their final form may be considered somewhat belated, no serious inconvenience results therefrom, as the wise precaution is taken of issuing the various sections in pamphlet form as soon as possible after the end of the year to which they refer, an advance sheet of statistics being, moreover, issued usually with considerable rapidity. This is a procedure that might well be imitated with great advantage by a good many other nations, our own not excepted. HENRY LOUIS.

THE INSTITUTION OF MECHANICAL
ENGINEERS.

THE summer meeting of the Institution of Mechanical Engineers opened at Liverpool on Tuesday, July 27. The president, Mr. John A. F. Aspinall, and the council and members of the institution, were welcomed in the lecture hall of the Municipal Central Technical School by the Lord Mayor of Liverpool, Councillor H. Chaloner Dowdall, and the members of the Liverpool reception committee. The importance of Liverpool as an engineering centre secured an attendance of nearly 500 members, who participated in the excellent arrangements made regarding visits to works and excursions. The institution dinner was held in the Exchange Station Hotel on Tuesday evening, and the Lord Mayor and Lady Mayoress of

on

Liverpool received the visitors in the Town Hall Wednesday evening. Meetings were held for the reading and discussion of papers on Tuesday and Wednesday mornings in the Municipal Central Technical School. Brief extracts from these are subjoined.

Locomotives designed and built at Horwich were described in a paper by Mr. George Hughes, who is the chief mechanical engineer of the Lancashire and Yorkshire Railway. This company possesses 1517 locomotives, of which there are about 1100 in daily use. When the works at Horwich were opened, Mr. Aspinall, president of the institution, and at that time chief mechanical engineer, resolved to introduce standardisation and, wherever possible, interchangeability. Joy's valve gear was adopted, as it was found that the mileage between repairs was greater, and also that there was a slight economy in coal per engine-mile.

Among other types of locomotives described it is of interest to note six engines which were fitted in 1902 with Druitt-Halpin thermal storage tanks. Where stopping places are frequent on rising gradients there is distinct economy. Certain tests carried out between Salford and Accrington resulted in a saving of one ton of water, and under similar conditions elsewhere the saving was 12 per cent. On other sections of the line, which are not so favourable, the all-round economy of these engines is brought down to 4 per cent.

A four-cylinder passenger and express goods engine, built to the author's designs in June, 1908, is also of interest. Absolutely perfect balancing could have been achieved without the aid of balance weights if the crank angles, the disposition of the cylinders, and the weights of the reciprocating parts had been arranged to neutralise amongst themselves the reciprocating disturbing forces; then, by balancing the revolving masses, the variations of rail load and the horizontal swaying couple would have disappeared. Excepting for a slight vertical component produced by the obliquity of the connecting-rod, the engine would then have been perfectly balanced. This arrangement, known as the Yarrow-Schlick-Tweedy system, would have involved an independent set of valve gear for each cylinder. Actually, the cranks were arranged in pairs at about 180° apart respectively, and the reciprocating masses, being made equal in weight, balance each other. The masses of the connecting-rods were divided between the rotating and reciprocating masses as suggested by Prof. Dalby, and the revolving masses were balanced by revolving balance weights. This engine is a very steady and smooth-running machine.

The discussion centred round the important questions of boiler deterioration, corrosion, and priming. Mr. Hurry Riches expressed the opinion that the best way of avoiding troubles due to the nature of the feed-water is to remove the impurities before feeding into the boiler; it is, however, inadvisable to reduce the hardness of feedwater below 6°.

A paper on reinforced concrete was contributed by Mr. Arthur C. Auden, of the firm of Messrs. William Cubitt and Co. Reinforced concrete is by no means a new thing; it has passed the experimental stage, as is evidenced by important structures erected in London in 1889, and still in use. On the Continent equally large structures exist which are now twenty-five years old, and have never been strengthened or patched. Failures have occurred through bad design or workmanship, but the proportion of these is small. The cost of the proposed structure is affected by the cost of its constituents, and these in turn by the cost of freight and carriage. Hence the author briefly classifies the materials, and gives useful hints on the properties of each.

For aggregates, the eastern counties' flint is often the only stone available locally. Good, tough concrete can be made with this, but is untrustworthy for fire-resisting purposes, owing to its tendency to crack and "fly" under heat. This tendency can be much reduced by first crushing all the stones. The same remarks apply to limestone, a material which is not more fire-resisting after being broken. As it is apt to disintegrate to powder under the action of heat, it is inadvisable to use this material where fire-resistance is an important consideration. Limestone always requires washing before use to get rid of the fine dust which covers it and prevents the cement properly

bonding with it. Sandstone, as a rule, is too soft, porous, and absorbent for use in reinforced-concrete work. It may be safely used if it will stand about 1 tons per square inch under a crushing test, and also if the difference in weight when clean and dry, and after being two days under water, does not exceed 8 per cent. Quartzite stone is fairly good. if not too soft and open in texture, in which case the same precautions apply as for sandstone. It should be noted that the test pieces for crushing tests should have an area of at least 10 inches or 12 inches.

With reference to artificially produced aggregates, broken earthenware and stoneware from the Potteries district make a good aggregate, but these must be unglazed, as the glaze prevents the proper adhesion of the cement. Burnt clay and gault may be used provided they are tough and hard, and do not soften or crumble after being left in water for two or three days. In general, broken bricks are not a good aggregate for reinforced concrete. They may be employed safely if hard and close in texture and free from mortar. Coke-breeze is cheap and readily obtained, but cannot be regarded as being really fire-proof. The effect of any sulphur present must also be considered. Ashes and clinkers may be used. In the case of ashes, only those which will float in water and are of uniform colour and texture, as well as being quite free from coal and dirt, should be used. Really hard and clean clinker alone is serviceable. In both of these sulphur must be considered. Slag from blast furnaces and cupolas makes a good aggregate if hard, tough, and free from dust; any sulphur present must be noted.

Sulphur is apt to attack the reinforcing steel with disastrous results. The maximum allowable percentage of sulphur in reinforced concrete aggregates is now being made the subject of experiments, and it is hoped an authoritative statement will soon be made. In the form of a sulphate sulphur is practically harmless, but is very deleterious if in the form of a sulphide.

It is of importance that no free lime be present in artificial aggregates; carbonate of lime is practically harmless. Washing and exposure to the air and sun will do much to convert sulphides into sulphates and free lime into carbonates. Good and accessible aggregates are often condemned because no discretion is exercised as to the form in which lime and sulphur occur.

A certain amount of sand is absolutely necessary in concrete, and no other material is at present known which can be substituted for it. Generally speaking, the better a sand is for moulding purposes the worse it is for reinforced concrete. Dirt in the form of slime, mud, or vegetable refuse is bad, but a little loam, enough to soil the fingers, but not enough to cause the sand to adhere to them, is no detriment. Small particles or nodules of clay do not appear to affect the strength of the concrete, but it is better to avoid them if possible. It is not good practice to use the stone aggregate, and its smalls and dust, together with some sand, upon the chance that they will be in proper proportion, and that the voids and spaces will be properly filled. Such a practice should not be allowed in reinforced work, where absolute homogeneity is so essential.

With reference to cement, any user is safe if he insists that his cement shall pass the British standard specification in every detail, and purchases from a trustworthy maker. It should not be one of the many mixtures imported into this country as cement, which do not deserve the name, and are costly at any price.

Methods of inserting the reinforcement in beams, slabs, columns, &c., together with hints on erecting various structures, take up the remainder of this valuable paper.

Prof. Unwin spoke of reinforced concrete construction as demanding excellent execution and supervision to be successful. In regard to formulæ of the empirical class. largely employed in this subject, the range of experimental work should rule the trustworthiness of the formulæ. Much of the present methods of design is based on guesswork. He took the opportunity of urging the necessity for more extended experiments.

In presenting his paper on the advance of marine engineering in the early twentieth century, Mr. Arthur J.

Maginnis naturally devotes a great deal of his space to the marine steam turbine. While the use of turbines has produced practically no advance or improvement in fr consumption since 1901, still, an advance has to be recorded in that a greater speed has been attained. During the past eight years experience has shown the trustworth ness of the Parsons turbine machinery. Notwithstanding that there are now more than seventy steamers continuously plying to and fro, no sailing schedules have been upset by a failure of machinery, nor has a turbine steamer ever had to be towed into port. The author has ne hesitation in stating that rotary machinery must eventually replace the present system in cargo steamers as well as in liners:

Combined systems of reciprocating and turbine machinery were referred to, but the author does not think that an extensive adoption of this system will be made. In evidence of the saving in weight in the boilers where turbines are installed, owing to the lower steam pressure which may be used, the author states that in the case of the Lusitania and Mauretania the saving in weight on the boilers alone is about 120 tons over and above that which would have been required if triple or quadruple piston engines had been used.

The author gives a summary of the results attained by marine engineering to date as follows:-vessels of close upon 800 feet length and more than 38,000 tons dis placement are being propelled across the Atlantic at an average speed of 25 knots by turbine machinery working up to about 70,000 horse-power, having a consump tion of upwards of 1000 tons per day. Similar results have been obtained in the turbine-propelled warship Indomitable, of more than 40,000 horse-power, and maintained across the Atlantic with water-tube boilers.

The electrical operation of textile factories formed the subject of a paper by Mr. Herbert W. Wilson. The principal advantage claimed lies in the fact that a much greater steadiness of drive can be obtained, with consequent higher average speed and increased output. Slight variations in speed above that corresponding to the maximum safe tension breaks the threads, and unless absolutely constant speed can be obtained, it is necessary to allow a margin of safety and to run at a speed materially below the breaking point. In one case in Lancashire, with two mills under the same management and of about the same size, and working under the same general conditions, the results obtained from the electrically driven factory have been distinctly superior to those from the mechanically driven one. The improvement in the quality of yarn was so noticeable that the output from the electrically driven mill fetched a distinctly better price than that from the other factory, the increase being stated at about 2 per cent. As regards increase in production, mills in this country which have adopted electrical driving may be estimated as showing an improvement of 5 per

cent.

A paper on the indicating of gas engines was contributed by Prof. F. W. Burstall, of Birmingham University. The Standards Committee of the Institution of Civil Engineers expressed the opinion in their 1906 report that the indicating of gas engines was open to very much greater errors than was the case with steam engines, and this matter has been considered by the Research Committee of the Institution of Mechanical Engineers. In the tests undertaken by the author, two indicators were used simultaneously, one of the ordinary string type and the other an optical indicator. A Premier gas engine was used having a cylinder 16 inches in diameter by 24 inches stroke, running at 165 revolutions per minute. The only variation in the four tests recorded was the amount of gas admitted, the mean pressure varying from 5 kg. per cm. up to about 7 kg. per cm.2

The string indicator emploved was of the Crosby type, selected for these tests by the Crosby Company. Before and after each set of trials the indicator was tested for backlash and friction, and the spring also calibrated. The backlash was in all cases negligible, and the friction amounted to less than 1 lb. with a spring having a scale of 400 lb. per square inch. The optic indicator was lent by Prof. Hopkinson, and was calibrated at the University. Both indicators were mounted on a branch piece con

nected to the engine cylinder, and the indicator diagrams were taken simultaneously. The indicator barrel of the Crosby indicator was rotated by a phosphor-bronze stranded wire wound round the barrel and led to a bellcrank lever. The bell-crank lever was driven by a steel wire attached to the usual lever driven by the engine piston. A very heavy spring, in which a compression of 400 lb. produced a contraction of 2 inches, controlled the bellcrank lever. The optic indicator was also driven by means of a phosphor-bronze stranded wire.

The mean diagrams were prepared from no fewer than twelve individual diagrams, each being divided by the method of ordinates, and the heights read by an accurate steel rule. With care it was possible to read the Crosby diagrams to an accuracy of half of 1 per cent. The optic indicator diagrams could readily be measured to the same order of accuracy. The diagrams were plotted on squared paper, and superposed one on the other, so as to exhibit whatever differences there were between the indicators. Speaking generally, the compression curves are coincident. The maximum pressures practically agree in two of the tests; in a third, the Crosby indicator gave the higher initial pressure, and in the fourth the Hopkinson gave the higher. Down the expansion line the two indicators agree for the third of the stroke. After that the Hopkinson indicator gave a persistently higher expansion line, the difference between the two lines being higher than the probable experimental error of the measurements. The effect of this difference is to make the Hopkinson indicator give about 3 per cent. higher mean pressure than the Crosby.

In the Hopkinson indicator the spring was in the form of a flat bar rigidly fixed at the ends and loaded in the centre; the central deflection of this beam is a direct measure of the pressure on the piston. During calibration with dead weights, from which the scale of the spring is obtained, the ends of the bar may be assumed to be absolutely fixed, but when the indicator is in use it is possible that there is a slight slip in the bar through the screws which restrain it. The effect of this would be to prevent the pressure falling so rapidly in this indicator as in the Crosby indicator. The author believes that this explanation is more likely to be correct than that the effect is due to inertia or friction, and is inclined to prefer the results obtained from the Crosby indicator.

While the results of this comparison do not offer an absolute proof of the accuracy of either indicator, there is still strong evidence that both give results very close to the truth. The indicators are of entirely different types, one multiplying the indicator piston movement by six, the other by about 120, a very similar multiplication being the case with the rotation of the drum and the mirror. In the optic mirror inertia is certainly negligible. That the two give results to within 3 per cent. on the mean pressure, and very nearly the same figures for the initial pressure, is good presumptive evidence that, when either indicator is used with the precautions regarding driving described, the results so obtained are at least as accurate as any other measurement which can be made in engine testing. Unless these precautions are taken, the results can only be regarded as affording a clue to the valve setting, and give no trustworthy figures as to the power developed in the engine cylinder.

The council of the institution has issued the conditions under which the second award of the water arbitration prize will be made in 1910. The prize will have a value of about 3ol., and will be awarded to the author of the best original paper dealing with any branch of the mechanics of the supply or distribution of water. The latest date for sending in papers will be September 1,

Mr. Haldane said he had made up his mind that there could be no real progress unless we proceed scientifically and in order-that is to say, unless we are perfectly clear about what we want, as to the structure of the machines which will be used for the purposes in view, and the production of them in a way which should be at least effective. The first thing done was to ask the Committee of Imperial Defence to investigate this question and to discuss it with the technical subcommittee. The report was to the effect that the class of machines must be divided into three heads-rigid dirigibles, non-rigid dirigibles, and aeroplanes. For naval purposes the rigid dirigible is probably the only instrument of the kind which is of real value, at any rate in the present state of knowledge. For the army the rigid dirigible has certain disadvantages. It is more difficult to work, to bring back, and to bring to rest. It is more difficult for the army than for the navy. The non-rigid dirigible is the best for army purposes. The aeroplane may become available for army purposes, but at present it has certain defects. It will have to rise much higher before it can be a safe instrument for reconnoitring. But M. Blériot's splendid feat in crossing the Channel and the successes achieved in the United States point to a time when the aeroplane may be an instrument capable of achieving great results.

To the navy has been assigned the duty of investigating, with the view of constructing, the rigid dirigible, the ship of the Count Zeppelin type. To the army has been assigned the duty of experimenting with the non-rigid dirigible, the machine of varying type, and also with aeroplanes.

at

on

To make their study of aviation scientific, Mr. Haldane said the Prime Minister constituted the advisory committee, under Lord Rayleigh's presidency, on which there is some of the best scientific brains in the world. Continuous work has been going on at the National Physical Laboratory. Meetings have been held there and Aldershot and the War Office. The committee is to advise, its purpose is to scrutinise inventions submitted in the course of the work of the departments concerned, and to conduct systematic experiments. In a few days the first report of this committee will be made public. The committee, said Mr. Haldane, has such men as Lord Rayleigh and Dr. Glazebrook on it, and such men the practical side as Mr. Lanchester and Mr. Mallock, and others like Prof. Petavel and Dr. Shaw, and also such high authorities on the army and navy side as Major-General Hadden and Rear-Admiral Bacon, and is well furnished from the various points of view. This committee has been at work, and the first thing it has done is to determine the general question which should be studied. There have been various memoranda by the experts on stability, screw propellers, wind structures, petrol motors, and a very difficult thing which has arisen in connection with balloons, the accumulation of electrostatic charges. Everyone knows what a peril electricity is in the air. Then the committee has mapped out the general field of its work. There are certain general questions in aerodynamics, questions specially relating to aeroplanes, such as the mathematical investigations of stability, the effect of rudder action, gusts of wind, and half a dozen other things which I need not enumerate. There are propeller experiments; there are questions relating to these motors which have to be of special construction for air work, general questions relating to airships, and still more general questions relating to meteorology.

The committee has entered into communication with the Aeronautical Society, the Aërial Club, and the Aëro League. The design is to afford assistance to private inventors wherever this can properly be done, because progress in this matter will be, not merely a Government, but a national matter. The Admiralty is concentrating, under Admiral Bacon, Director of Naval Ordnance, on the building of a rigid dirigible of the very largest type, at least the size of the Zeppelin. That is being built at Barrow-in-Furness by Messrs. Vickers. The combination of experts and practical men may give us a practical result some time next spring. The War Office is reorganising its factory at Aldershot. The instruction, which is at present given to balloonists under the superintendence of Colonel Capper, is being separated from construc

tion, and at present preparations are being made for the construction of a shed which will take in the largest size of dirigible. The Admiralty has in prospect one great rigid dirigible, the War Office has three, and besides those we have our balloons for war purposes. At the present time we have certain aeroplanes, and the prospect of two new aeroplanes which are to be presented for experimental purposes, and may hereafter be acquired. That is the actual position of things.

IMPROVEMENTS IN PRODUCTION AND APPLICATION OF GUNCOTTON AND NITROGLYCERINE.1

II.

IN the year 1846 Schönbein discovered guncotton. In the year 1886, that is, forty years later, the French chemist Vieille invented his smokeless powder for military purposes. This explosive, which was primarily designed for use in the small calibre Lebel rifle, consisted essentially of guncotton, and the secret of its success lay in the fact that Vieille so altered its physical state that its rate of combustion, when confined, was under complete control. This condition was arrived at by treating the fibrous guncotton with suitable solvents which entirely destroyed the fibre and converted it into a colloidal, horny substance quite devoid of all porosity. The gelatinised guncotton resulting from this treatment burnt, when ignited, from the surface inwards, and by varying the surface area any required rate of combustion could be obtained. The use of smokeless powders manufactured in this way was very soon extended to all natures of ordnance.

The next step in the development of smokeless powders was the combination of nitroglycerine with nitrocellulose. The first powder of this type was the " ballistite " of Alfred Nobel, patented by him in the year 1888. The original ballistite was composed of equal parts of nitroglycerine and of soluble nitrocellulose, a variety of guncotton soluble in nitroglycerine, and no solvent was therefore required in its preparation, although a certain proportion of camphor was used to promote the solution of the nitrocellulose. Another form of nitroglycerine-nitrocellulose explosive is the British service powder, cordite, which originally consisted of nitroglycerine, 58 parts, guncotton, insoluble in nitroglycerine, 37 parts, and mineral jelly, a product of the distillation of crude petroleum, 5 parts. To effect the gelatinisation of the guncotton, the solvent acetone, obtained indirectly from the destructive distillation of wood, is employed. The result of subjecting nitrocellulose in suitable machines to the action of nitroglycerine or of solvents, of which there are several suitable ones besides acetone, is to destroy its fibre and convert it into a gelatinous mass, in which condition it can be formed into any desired shape. Where solvents are used to produce this result they remain in the mass during subsequent operations, and are finally driven off by means of heat. The resulting products, somewhat incorrectly termed " powders," which are manufactured in a variety of forms, such as grains and flakes of different shapes, ribbons or strips, solid cords, tubes, &c., vary in consistence with the quantity of nitroglycerine they contain. The more nitroglycerine present the softer the powder, pure nitrocellulose powders being hard to brittle

ness.

For practical purposes modern smokeless powders are of two types :

(1) These consisting entirely of nitrocellulose, and termed "nitrocellulose powders."

66

(2) Those consisting of a mixture of nitrocellulose and nitroglycerine, known as nitroglycerine powders.' Opinions differ somewhat as to the relative merits of these two types; in this country the latter type is preferred. Their characteristic features are, briefly, as follows:

A nitroglycerine powder is more powerful than a nitrocellulose powder, and the more nitroglycerine present the more powerful the explosive. Therefore, for equal ballistics, a smaller charge of the former than of the latter is required, and, consequently, the chamber capacity and 1 Discourse delivered at the Royal Institution on Friday, January 29, by Sir Frederic L. Nathan, R.A. Continued from p. 147.

the size and weight of the breech mechanism are reduced; on the other hand, the higher the proportion of nitroglycerine the higher is the temperature of combustion and the greater the erosive effects on the surface of the bore of the gun.

The presence of nitroglycerine in an explosive allows of the more easy and rapid elimination of the solvent used in manufacture and of moisture, a small quantity of which is always present in nitroglycerine and guncotton. The sooner this is attained the better, because the longer the time that the powder is being heated in order to dry it, the more likely is its chemical stability to be affected. Moreover, it is a well-established fact that with nitrocellulose powders it is impossible to remove the volatile matter with anything like the same completeness as can be done in the case of nitroglycerine powders. The consequence is that the slow evaporation from nitrocellulose powders of the residual volatile matter which takes place in store tends to produce changes in their physical character and renders them in course of time liable to alter in ballistic properties, and even to develop dangerous pressures in the gun.

Nitroglycerine powders are cheaper than nitrocellulose powders, weight for weight, and even more so for equal ballistic effects.

The original cordite, the manufacture of which commenced in 1890, contained a high proportion of nitroglycerine, 58 per cent., and the erosion produced, especially in large guns, was considerable. This led to experiments being carried out at Waltham Abbey with the view of the production of a less erosive explosive, and the final result was the introduction into the service, in 1901, of a modified cordite known as "cordite M.D.," in which the percentage of nitroglycerine is reduced to 30 per cent., so that the composition becomes :-nitroglycerine, 30 per cent.; guncotton, 65 per cent.; and mineral jelly, 5 per

An inspection of these figures shows that the alteration in proportions of the explosive ingredients results in a decrease in the heat of explosion of about 16 per cent., and an increase in the volume of gases of about 5 per cent., whilst there is a decrease of 289° C. in the temperature of explosion.

As would therefore be expected, the erosion produced by cordite M.D. is very much less than that produced by the original cordite for the same ballistics, and is certainly not greater, if as great, as that produced by the best forms of nitrocellulose explosives.

Although of minor importance to smokelessness, flamelessness is a desirable quality for propulsive explosives to possess. In this respect cordite M.D. is superior to cordite in the case of rifles and machine guns; unfortunately, a suitable ingredient has not yet been discovered which will render smokeless powders flameless in large guns.

A third ingredient in both natures of cordite, viz. mineral jelly, although present in a comparatively small proportion, is a very important constituent.

Cordite in the advanced experimental stage consisted of nitroglycerine and guncotton alone, and as their combustion produced no solid residue of any kind, the surface of the bore of the magazine rifle in which the early experiments took place was not fouled in any way. The result was that the cupro-nickel coated bullets, propelled in succession at high velocity through a clean barrel, deposited some of the cupro-nickel in the bore. In order to prevent this a number of substances were incorporated with the nitroglycerine and guncotton, with the object

of producing a deposit in the bore, which it was hoped would get rid of the difficulty of metallic fouling. Of all these various substances the one which appeared to answer the purpose most satisfactorily was refined vaseline, and this material became the third ingredient of cordite as eventually introduced into the British service. When the manufacture was commenced on a large scale, vaseline, which is the proprietary name of one of the refined products of the distillation of petroleum, was replaced by mineral jelly, the same material, but in a cruder form.

The original object with which mineral jelly was introduced was of no importance when cordite was substituted for the black and brown powders used in large guns, but in order to have but one nature of smokeless powder in the service mineral jelly was added to all cordite, whether for use in small arms or artillery. Subsequent experience has demonstrated how very fortunate was the selection of this material for rifle cordite and the extension of its use to all sizes of cordite.

Mineral jelly is one of the best ingredients it is possible to have in smokeless powders from the point of view of their chemical stability. This important fact, not recognised originally, was brought out in the following way. In order to facilitate the explosion of cordite in blank ammunition for the rifle it was cut into very thin flakes, and the non-explosive mineral jelly was omitted from its composition. After a comparatively short storage in a hot climate the stability of the smokeless blank, as it was called, was found to have suffered seriously, whereas the stability of normal cordite containing mineral jelly was not appreciably affected. These facts led to a thorough investigation at Waltham Abbey of the action of mineral jelly in preserving the stability of cordite, and it was discovered that mineral jelly contained constituents which had the valuable property of combining with the decomposition products (the result of prolonged storage of cordite at high temperatures) to form stable bodies, thus removing these decomposition products, which undoubtedly exert a deteriorating influence on the cordite from their sphere of action.

When Abel was engaged on his researches in connection with the production and properties of guncotton, it was obvious to him that some test of a chemical nature was required in order to ascertain whether or not the finished guncotton had been thoroughly purified in manufacture. It will be remembered that accidents occurred in the early days of its production because this purification had not been carried sufficiently far. The test which he devised was based on the principle that if guncotton be subjected to an elevated temperature, traces of oxides of nitrogen will be given off, and will reveal their presence by acting on a suitable reagent.

The test is carried out by heating guncotton in a testtube placed in a water bath, and suspending over it a strip of moistened filter paper impregnated with potassium iodide and starch. If the purification of the guncotton has not been sufficient, the discoloration of the test paper takes place early; as the result of experience Abel fixed a time before which no reaction should take place. This test, known as the Abel heat test, is a test for the purity of guncotton and of nitroglycerine, and of freshly made explosives containing either one or both of these ingredients. For this purpose no test has yet been devised which equals it. But it was never intended to be, and is not, a quantitative test, and is therefore only a rough guide, though a very useful one, as to the stability of an explosive which has been in store for more or less prolonged periods, or under more or less adverse conditions.

Smokeless powders of the types dealt with are all subject to deterioration, and there is very little doubt that this deterioration is for any given explosive a function of the temperature of storage. The higher the temperature the more rapid the deterioration.

a test

The necessity, therefore, of some quantitative test which would enable a judgment to be formed as to the extent of deterioration suffered by any given sample of cordite is obviously of great importance, because such would afford the means of determining how much longer it would be safe to store any given batch of cartridges or lot of cordite at any given temperature. Any such test

must be a heating test, and it must be possible to correlate the temperature and duration of the test with any given temperature and duration of storage. The rate of deterioration as a function of the temperature was determined by Dr. Will for guncotton, and later by Dr. Robertson at Waltham Abbey for nitroglycerine. From these and other experiments carried out at Waltham Abbey, a factor of increase in rate of deterioration of cordite with increase of temperature was deduced. This factor having been determined, what is known as the silvered vessel test" was worked out at the Royal Gunpowder Factory. In this test, of which the details will be described presently, cordite is heated in a specially designed vessel at 80° C., a temperature not too far removed from those to be met with when cordite is stored under the worst service conditions, and the number of hours' heating at this temperature any given sample will stand before it shows signs of active decomposition are ascertained. Then, by means of an equation, containing the factor connecting rate of increase of deterioration with rise in temperature, a calculation can be made converting the hours of heating at 80° C. the sample withstood to years and fractions of a year it would stand at any given temperature of storage, and therefore a knowledge is obtained of how much longer it would be safe to store this cordite at any given tempera

ture.

This test was applied to a considerable number of samples of known age and thermal history. From these data, and knowing the number of hours at 80° C. that newly made cordite of good stability will stand before showing signs of decomposition, the number of hours that the different samples should stand the test were calculated. When the samples were actually tested, the number of hours' heating at 80° C. they withstood were in close agreement with the number of hours it was calculated they should stand.

The form of vessel in which the heating is carried out is the well-known vacuum vessel of Sir James Dewar. A glass bulb silvered externally is enclosed in an outer bulb silvered internally. The space between the two is highly evacuated for the purpose of limiting the dissipation of any heat evolved by exothermic changes on the one hand, and on the other for the purpose of minimising the effect of accidental slight changes in temperature of environment.

In the centre of the inner bulb is situated the bulb of a thermometer, the stem of which passes through a cork in the neck of the vessel. A side tube is attached for the purpose of making observations on the colour of the gases evolved. For heating the vessel a bath is provided with cylinders closed at the bottom, and wide enough to admit the vessel to such a depth as the side tube will permit. The bath is surrounded by insulating materials. vessels are packed in the cylinders with wool yarn, and the tops of the cylinders are closed with felt discs to exclude draughts.

The

The bath is fitted with a gas regulator or other means for securing that the temperature of the explosive is kept

constant.

The cordite is coarsely ground, and 50 grams are used. Readings of the thermometer are taken at intervals, and the time is noted when a rise of 2° C. in the temperature of the explosive above the temperature of 80° C. occurs. At the same time, visual observations are made as to the colour of the column of gas in the side tube, since it is found that, previous to the rise in temperature occurring, orange-coloured fumes of nitric peroxide are evolved. When the temperature exceeds 82° C. the test is complete, and the flask is withdrawn. The number of hours which have elapsed since the start of the test is the measure of the stability of the cordite.

Until about sixty years ago, the only explosive known for all purposes was gunpowder. With the discovery of guncotton and nitroglycerine, gunpowder was gradually replaced by them for blasting purposes. In their early days the two explosives were used singly, guncotton as guncotton, nitroglycerine-first of all alone-and then as dynamite. Later on the two were combined as blasting gelatine and explosives of a similar nature, but it was quite forty years after their discovery before either became of practical use for propulsive purposes.