V.-Aid to Scientific and Professional Societies. At the outset of their work, the Advisory Council were able to prevent a number of important researches undetaken by scientific and professional societies from being abandoned through lack of funds. The total grants made to organisations of this kind in respect of 69 researches during the five years under review amounted to £68,816. The financial assistance afforded to societies tends, however, to diminish, since some of the researches are being transferred to research associations and some to the Department itself; in the latter case the society often continues to assist by appointing a Supervising Committee; an additional reason for the diminution of the grants to societies in the fact that the call on their funds for aid for researches intended solely to increase knowledge has been lessended on account of the grants made by the Department to individual workers. Amongst grants to scientific societies are those to the Institution of Civil Engineers for the valuable work they have carried out on the deterioration of sea structures, to the Silk Association for the researches on the degumming of silk, to the Institution of Gas Engineers for the work of Dr. Mellor on Gas Retorts, to the Institution of Chemistry for their work on laboratory and optical glass, to the Pianoforte Manufacturers' Association for Dr. Clay's researches on sounding boards, to the British Fire Prevention Committee, to the Royal Society for Sir William Pope's researches on colour sensitiser dyes, and to the Manchester Association of Engineers for Mr. Dempster Smith's researches on the hardening of tool steel. Valuable work has been done for the Institution of Metals by Dr. Bengough and his colleagues on the corrosion of condenser tubes. Grants have also been made to three educational institutions for work of urgent industrial importance to the Stoke-on-Trent Central School of Science and Technology for researches on hard porcelain; to Sheffield University for work on glass technology; and to the Imperial College of Science and Technology for researches in technical optics. In the case of two of the industries concerned in the work aided at these educational institutions, research associations have been formed, while the negotiations with the third for the formation of research association have reached an advanced stage. a VI.-Developments during the Year 1919-20. The body of the Report itself gives a detailed account of the activities of the Department during the year 1919-20, and include reference to the development of research associations, the organisations of national research, the expansion of research in and for the Empire, aided research, patents, and the publications of learned societies. ADDRESS TO THE ENGINEERING SECTION OF THE BRITISH ASSOCIATION. By Professor C. F. JENKIN, C.B.E., M.A., THE importance of research in all branches of industry is now becoming fully recognised. It is hardly necessary to point out the great possibilities of the Board of Scientific and Industrial Research, formed just before the war, or to lay stress on the attention which has been called to the need for research by events during the war. Probably in no branch of the Services was more research work done than in the Air Service, and the advances made in all directions in connection with flying were astonishing. My own work was confined to problems connected with materials of construction, and as a result of that work I have come to the conclusion that the time has come when the fundamental data on which the engineering theories of the strength and suitability of materials are based require thorough overhauling and revision. I believe that the present is a favourable time for this work, but I think that attention needs to be drawn to it, lest research work is all diverted to the problems which attract more attention, owing to their being in the forefront of the advancing engineering knowledge, and lest the necessary drudgery is shirked in favour of the more exciting new discoveries. It has been very remarkable how again and again in aeroplane engineering the problems to be solved have raised fundamental questions in the strength and properties of materials which had never been adequately solved. Some of these and some related to the physical properties of questions related to what may be termed theory, materials. I propose to-day to describe some of these problems, and to suggest the direction in which revision and extension of our fundamental theories and data are required and the lines on consider first one of the oldest materials of conwhich research should be undertaken. Let us struction-timber. Timber was of prime impority of this material which strikes us is that it is tance in aircraft construction. The first peculiaranistropic. Its grain may be used to locate three the grain, and tangentially across the grain. It principal axes-along the grain, radially across is curious that there do not appear to be generally recognised terms for these three fundamental directions. A very few tests are sufficient to show that its strength is enormously greater along the grain than across it. How, then, is an engineer to calculate the strength of a wooden member? There is no theory, in a form available for the engineer, by which the strength of members made of an anistropic material can be calculated. I fancy I may be told that such a theory is not required that experience shows that the ordinary theory is quite near enough. How utterly misleading such a statement is I will try to show by a few examples. Suppose a wooden tie or strut is cut from the tree obliquely so that the grain does not lie parallel to its length. In practice it is never possible to ensure that the grain is accurately parallel to the length of the member, and often the deviation is considerable. How much is the member weakened? This comparatively simple problem has been of immense importance in aeroplane construction, and, thanks to the researches made during the war, can be answered. The solution has thrown a flood of light on many failures which before were obscure. If the tensile strengths of a piece of timber are, say, 18,000 lb./sq. in. along the grain and 800 lb/sq. in. across it (radially or tangentially) and the shear strength is 900 lb./sq. in. along the grain-these figures correspond roughly with the strengths of silver spruce-then if a tensile stress be applied at any angle to the grain the components of that, 1 172 CHEMICAL. NEWS Address to Engineering Section-British Association [October 6, 1920 stress in the principal directions must not exceed the above strengths, or failure will occur. Thus we can draw curves limiting the stress at any angle to the grain, and similar curves may be drawn for compression stresses. These theoretical curves have been checked experimentally, and the results of the tests confirm them closely, except in one particular. The strengths at small inclination to the grain fall even faster that the theoretical curves would lead us to expect. The very rapid drop in strength for quite small deviations is most striking. Similar curves have been prepared for tensile and compressive stresses inclined in each of the three principal planes for spruce, ash, walnut, and mahogany, so that the strengths of these timbers to resist forces in any direction can now be estimated reasonably accurately. As a second example consider the strength of plywood. Plywood is the name given to wood built up of several thicknesses glued together with the grain in alternate thichnesses running along and across the plank. The result of this crossing of the grain is that the plywood has roughly equal strength along and across the plank. Plywood is generally built up of thin veneers, which are cut from the log by slicing them off as the log revolves in a lathe. Owing to the taper in the trunk of the tree and other irregularities in form, the grain in the veneer rarely runs parallel to the surface, but generally runs through the sheet at a more or less oblique angle. As a consequence, the strength of plywood is very variable, and tests show that it is not possible to rely on its having more than half the strength it would have if the grain in the veneers were not oblique. It is, therefore, obviously possible to improve the manufacture enormously by using veneers split off, following the grain, in place of the present sliced vencers. The superiority of split or riven wood over cut wood has been recognised for ages. I believe all ladders and ladder rungs are riven. Hurdles, hoops, and laths are other examples. Knees in ships are chosen so that the grain follows the required outline. Owing to the enormous difference in strength in timber along and across the grain, it is obviously important to get the grain in exactly the right direction to bear the loads it has to carry. The most perfect example I ever saw of building up a plywood structure to support all the loads on it was the frame of the German Schutte-Lanz airship, which was made entirely of wood. At the complex junctions of the various girders and ties the wood, which was built up of very thin veneers -hardly thicker than plane shavings-layers were put on most ingeniously in the direction of every stress. During the war I have had to reject numerous types of built-up struts intended for aeroplanes, because the grain of the wood was in the wrong direction to bear the load. The example showna McGruer strut--is one of the most elegant designs, using the grain correctly. Many of the tests applied to timber are wrong in theory and consequently misleading. For example, the common method of determining Young's modulus for timber is to measure the elastic deflection of a beam loaded in the middle and to calculate the modulus by the ordinary theory, neglecting the deflection due to shear, which is legitimate in isotropic materials; but in timber the shear modulus is very small-for example, in spruce it is only about one-sixtieth of Young's modulus-and consequently the shear deflection becomes quite appreciable, and the results obtained on test pieces of the common proportions lead to errors in the calculated Young's modulus of about 10 per cent. The lantern plates show three standard tests; the first is supposed to give the shearing strength of the timber, but these test pieces fail by tension across the grain-not by shearing. Professor Robertson has shown that the true shear strength of spruce is about three times as great as the textbook figures, and has designed a test which gives fairly reliable results. The second figure represents a test intended to give the mean strength across the grain, but the concentration of stress at the grooves is so great that such test pieces fail under less than half the proper load. This fact was shown in a striking manner by narrowing a sample of this shape to half its width, when it actually bore a greater total load-i.e., more than double the stress borne by the original sample. The third figure represents a test piece intended to measure the rather vague quality, "strength to resist splitting." The results actually depend on the tensile strength across the grain, on the elastic constants, and on the accidental position of the bottom of the groove relatively to the spring or autumn wood in the annular rings. Unless the theory is understood, rational tests cannot be devised. There are some valuable tropical timbers whose structure is far more complex than that of our ordinary northern woods. The grain in these timbers grows in alternating spirals-an arrangement which at first sight is almost incredible. The most striking example of this type of wood I have seen is the Indian "Poon." The sample on the table has been split in a series of tangential planes at varying distances from the centre of the tree, and it will be seen that the grain at one depth is growing in a right-hand spiral round the trunk; a little farther out it grows straight up the trunk; further out it grows in a left-hand spiral, and this is repeated again and again, with a pitch of about two inches. The timber is strong and probably well adapted for use in large pieces-it somewhat resembles plywood-but it is doubtful whether it is safe in small pieces. No theory is yet available for estimating its strength, and very elaborate tests would be needed to determine its reliability in all positions. I had to reject it for aeroplanes during the war for want of accurate knowledge of its properties. These examples show how necessary it is to have a theory for the strength of anistropic materials before we can either understand the causes of their failure or make full use of their properties or even test the rationally. The second material we shall consider is steel, and in dealing with it I do not wish to enter into any of the dozen or so burning questions which are so familiar to all metallurgists and engineers, but to call your attention to a few more fundamental questions. Steel is not strictly isotropicbut we may consider it to be so to-day. The first obvious question the engineer has to answer is, "What is its strength?" The usual tests give the NEWS Ultimate Strength, Yield Point, Elastic Limit, the Elongation, the Reduction of Area, and perhaps the Brinell and Izod figures. On which of these figures is the dimension of an engine part, which is being designed, to be based? If we choose the Ultimate Strength we must divide it by a large factor of safety-a factor of ignorance. If we choose the Yield Point we must remember that none of the higher-grade steels have any Yield Point, and the nominal Yield Point depends on the fancy of the tester. This entirely imaginary point cannot be used for accurate calculation except in a very few special cases. Can we base our calculation on the Elongationthe Reduction of Area- the Izod test? If we face the question honestly we realise that there is no known connection between the test results and the stress we can safely call on the steel to bear. The only connecting link is that cloak for our ignorance-the factor of safety. I feel confident that the only reliable property on which to base the strength of any engine part is the suitable Fatigue Limit. We have not yet reached the position of being able to specify this figure, but a considerable number of tests show that in a wide range of steels (though there are some unexplained exceptions) the Fatigue Limit for equal stresses is a little under half the Ultimate Strength, and is independent of the Elastic Limit and nominal Yield Point, so that the Ultimate Strength may be replaced as the most relable guide to true strength, with a factor -no longer of ignorance, but to give the fatigue limit-of a little over 2. If the Fatigue Limit is accepted as the only sound basis for strength calculation for engine parts, and it is difficult to find any valid objection to it, then it obvious that there is urgent need for extensive researches in fatigue, for the available data are most meagre. The work is laborious, for there is not one Fatigue Limit, but a continuous series, as the signs and magnitudes of the stresses change. Many problems in connection with fatigue are of great importance and need much fuller investigation than they have so far received-e.g., the effect of speed of testing; the effect of rest and heat treatment in restoring fatigued material; the effect of previous testing at higher or lower stresses on the apparent fatigue limit of a test piece. Some observers have found indications that the material may possibly be strengthened by subjecting it to an alternating stress below its fatigue limit, so that the results of fatigue tests may depend on whether the limit is approached by increasing the stress or by decreasing it. Improved methods of testing are also neededparticularly methods which will give the results quickly. Stromeyer's method of measuring the first rise of temperature, which indicates that the fatigue limit is passed, as the alternating load is gradually increased, is most promising; it certainly will not give the true fatigue limit in all cases, for it has been shown by Bairstow that with some ranges of stress a finite extension occurs at the beginning of a test and then ceases, under stresses lower than the fatigue limit. But the fatigue limit in that case would not be a safe guide, for finite changes of shape are not permissible in most machines, so that in that case also Stromeyer's test may be exactly what is a wanted. It can probably be simplified in detail and made practicable for commercial use. Better methods of testing in torsion are also urgently needed, none of those at present used being free from serious defects. Finally, there is fascinating field for physical research in investigating the internal mechanism of fatigue failure. Some most suggestive results have already been obtained, which extend the results obtained by Ewing. For members of structures which are only subjected to steady loads I suggest that the safe stress might be defined by limiting the corresponding permanent set to a small amountperhaps per cent or per cent. This principle has been tentatively adopted in some of the aircraft material specifications by specifying a Proof Load which must be sustained without a permanent extension of more than per cent. Whether this principle is suitable for all materials and how it will answer in practice remains to be proved by experience. It is at any rate a possible rational basis for determining the useful strength of a material under steady loads. The relation between the proof stress and the shape of the stress-strain diagram is shown in the lantern slide. The curve is the record of an actual test on a certain copper alloy. If a length AB corresponding to per cent elongation be set off along the base line and a line BP be drawn through the point B parallel to the elastic line, to cut the curve in P, then the stress at P is the stress which will give per cent permanent set. Though per cent may appear rather a large permanent set to allow it will be seen from the figure that it is less than the elastic elongation would have been at the same stress, and we do not usually find elastic elongations serious. As a commercial test the proof load is very easily applied. For this alloy the specified proof load is shown by the horizontal line so labelled. This load is to be applied and released, and the permanent extension is required by the specification to be less than per cent. This sample passes the test easily. On the figure the condition for complying with the specification is that the curve shall fall above Q. But the test does not require the curve to be determined. (To be continued.) THE VOLUMETRIC METHODS FOR ESTIMATING TIN. By J. G. F. DRUCE, M.Sc., A.I.C. IN the course of certain investigations on the double chlorides of tin it has been necessary to carry out many quantitative analyses for tin. Frequently the metal was estimated by a gravimetric process. In those cases where the tin was present in the stannous condition it was possible to estimate it by volumetric methods. A number of these processes have been tried in a systematic examination to find out the most suitable. Most volumetric methods were found to be satisfactory but it may not be generally realised that stannous tin can be most easily and accurately determined in acid solutions. The results obtained with the various oxidising solutions are given below. Potassium stannochloride K,SnCl,,2H,O was chosen as the most SnO+H,O+I,--SnO,+2HI. Muller (Bull. Soc. Chim., 1900, XXV., 1002), modified the method by dissolving the substance in hydrochloric acid, and then added Rochelle salt together with sodium carbonate. For this method, 5174 grms. of potassium stannochloride were dissolved in 50 CC. of dilute hydrochloric acid (one volume of concentrated acid to three of water) The solution was made up to 250 CC. with water containing one per cent of sodium carbonate dissolved in it in order to minimise oxidation caused by the presence of air. For each titration, 10 cc. of this solution (01008 normal) were mixed with an equal volume of a saturated solution of Rochelle salt, and finally 25 cc. of a saturated sodium bicarbonate solution were added. The solution was then alkaline to litmus, and was titrated with a decinormal solution of iodine, a little starch paste being added as indicator towards the end of the titration. Warming was quite unnecessary. It will be observed that the results are a little low. Ten cc. of stannous solution required : 10'50 10'55 10'55 10'45 0'05626 0'05984 0.05609 0'05984 0'05609 0'05984 0'05669 0'05984 Estimation of tin by titrating stannous solutions with standard iodine solution in the presence of hydrochloric acid has been successfully employed by S. W. Young (J. Amer. Chem. Soc., 1897, xix., 515 and 807), and by Ibottson and Brearley (CHEMICAL NEWS, 1901, lxxxiv., 167) who obtained very satisfactory results. The hydrochloric acid solution of the tin salt was titrated with standard iodine solution until the end point was almost reached, when a little starch paste was added. The following reaction took place : 2SnCl,+2I,=SnCl+SnI. In order to obtain a paste which would give a "homogeneous" appearance to the blue-black starch iodide, the starch (5 grms.) was first ground in a mortar with 10 cc. of water and then poured into 200 cc. of boiling water, and ebullition continued for five minutes. 10 cc. of the above potassium stannochloride solution required : Using a fresh solution containing 20 84 grms. K,SnCI,,2H,O per litre (i.e., 0' 1012 normal). 4 5 6 10'05 This seemed to be the most convenient process for estimating tin by volumetric means. It gave satisfactory results for stannouschloride, the inorganic stannochlorides (CHEMICAL NEWS, 1918, cxvii., 193), and the stannochlorides of the aliphatic amines (ibid., 1919, cxviii., 1), but was less suitable for most of the aromatic amine compounds owing to the tendency of solutions of these to darken during the titration and thus obscure the end point. Lowenthal (/. pr. Chem., 1858, lxxvi., 484) has described a method whereby stannous chloride was treated with excess of ferric chloride, which becomes reduced to ferrous chloride, which was titrated with potassium permanganate solution. 2 FeCl,+SnCl, SnCl,+2FeCl,. His method has now been repeated but the results are not so good as those above. It is not very reliable since the end point is somewhat uncertain. Using decinormal potassium permanganate solution, 10 CC. of the second tin solution required : A satisfactory method of estimating tin by means of permanganate has, however, been devised in which the hydrochloric acid present during titration was kept at a minimum by dissolving the stannous salt in dilute sulphuric acid and titrating the solution as soon as possible after it had been prepared. The end point was quite sharp and the results were good as the following shows. 5 205 grms. potassium stannochloride were dissolved in 50 cc. of dilute sulphuric acid, and the solution was made up to 250 cc., with one per cent sodium carbonate solution. 10 per cent of this solution then required: NEWS . 8 Cr,O,+2SnCl2+14HCl = 2KCl+2SnCl,+Cr,Cl,+7H2O Stannous chloride was found to act upon the ndicator to produce sulphanilic acid stannihloride (NH,.CH,SO,H), H,SnCl (see CHEMIAL NEWS, 1919, cxix., 74), when heated for some ime in dilute hydrochloric acid. The salt began › darken at about 265-270° C., and it yielded ulphanilic acid after the removal of tin as sulhide. Hence the indicator is probably azobenene-pp'-disulphuric acid, CERTAIN of the factors which determine the perneability of rubber to gases have been investigated and the relative rates of penetration of a number of gases determined. The major findings may be summarised as follows: 1. The permeability of rubber compounds varies with the composition as would be expected. The aging of rubber films is accompanied by a decrease in permeability. A similar decrease may e affected by over-vulcanisation. The rubber which shows a very low permeability for these reasons is usually very much deterioated and frequently brittle so that it is a disadvantage from the standpoint of gas lighters. 2. The permeability to any gas is found to be directly proportional to its partial pressure provided the total pressure is constant. The variation of permeability with total pressure depends on the thickness of the rubber, the way in which it is supported, &c. 7. The permeability of rubber to water vapour is high-approximating 50 times the permeability to hydrogen. The value not having been determined with any precision is not included in the table above.-Journ. Franklin Inst., August, 1920. MINERAL RESOURCES OF THE WORLD. Magnesite.* *Pamphlet issued by the Imperial Mineral Resources Bureau, entitled "The Mineral Industry of the British Empire and Foreign Countries." Magnesite is essentially magnesium carbonate; it may be spathic and comparatively coarse in its crystallisation, containing a considerable percentage of ferrous carbonate, or it may be chalk-like compact, or crypto crystalline in texture. as Magnesite is widely distributed over the surface of the earth. Austrian magnesite is spathic, and contains ferrous carbonate, and is known breunnerite. Chalk-like magnesite, which is the commercially important, occurs in India, AustraAnother lia, Greece, Italy, and California. variety of magnesium carbonate, hydromagnesite, is a distinct mineral species, with a chemical composition and physical properties essentially different from those of magnesite. Deposits are found in British Columbia and Spain. Table I. gives the approximate analyses of different forms of magnesite. 3. The permeability to hydrogen is inversely proportional to the thickness of the rubber. other gas was tested in this respect. 4. The specific permeability to hydrogen at 25° C. of vulcanised rubber similar to the grade known as dental dam is about 20× 10-6 cc. per minute. This value varies somewhat with the age and chemical characteristics of the rubber. Magnesite is utilised chiefly as a source of magnesia, which is obtained from it by calcination in kilns (1) When heated to a temperature not |