It has been found that bentonite is pretty widely distributed in the upper sections of the Cretaceous rocks which cover the greater part of the Prairie Provinces, and that many of the sandy or shaly beds in this series of rocks are bentonite in varying degree. The sticky "gumbo" soils of the west owe their character to the presence in them of bentonite, which swells up when wet and forms a pasty mass that makes the prairie roads at wet seasons almost impassable. The University of Alberta is at the present time conducting experimental work in the improvement of highways by counteracting the evil effects of bentonite.

The bentonite beds of the Canadian West range in thickness from a few inches to .s much as 14 feet. The latter thickness is exceptional, but several deposits measuring 6 to 8 feet are known. The deposits which probably offer the best possibilities for immediate development occur in southern Saskatchewan and southern British Columbia, where beds are found quite close to railway transportation. No production from any of the deposits has yet been made. owing to the lack of a market. In th United States, however, two companies are now marketing bentonite; one of these produces only crude clay, and the other both crude and powdered material. These companies report that the market is small us yet, but is growing steadily each year, and that an encouraging interest is being manifested in the possibilities of the clay by the most varied industries. In 1921, the Mines Branch arranged the shipment of 5 tons of Alberta bentonite to the Imperial Mineral Resources Bureau, 2, Queen Anne's Gate Buildings, London, for research purposes and for distribution in British industries. It is understood that the Bureau still has a quantity of the clay on hand.

Bentonite in the dry state looks much like any other clay. It is only when it is wetted that its peculiar physical properties become apparent. The name bentonite was given about twenty-five years ago to the clay from a bed that occurs in rocks of Upper Cretaceous age (Fort Benton series) in Wyoming, and attention was drawn to the material on account of its excessively sticky nature and its high water absorption. It possessed, in short, the physical attributes of what are now called colloids.' Briefly, bentonite is a clay substance, the individual particles of which are so small that they are invisible even under the

highest powered microscope: they are not crystalline like ordinary clay particles, and when mixed with an excess of water form what are called "hydro-sols"-that is, permanent suspensions, or dispersions, from which the clay will not settle out, even if allowed to stand indefinitely. By virtue of the relatively enormous surface presented by the particles, bentonite possesses great absorptive power, and the dry clay will absorb many times its volume of water, swelling up to a sticky, jelly-like mass. The particles have the property of absorbing chemical compounds also, such as salts, dyes, etc., out of solution, and of effecting base exchange between salts of the alkali and alkali earth groups.

These properties have suggested a very wide field of usefulness for bentonite in industry. Some of the more important of these uses are: as an absorbent of dyes and colours; to increase the strength of cements and plasters: as a suspending agent for enamels and glazes in the metal enamelling and ceramic industries; as a bonding agent in crucible bodies, electrical and chemical porcelain and abrasive wheels; as a dewatering agent for petroleum, gasoline and oils generally, as well as for air and gases; as a mordent and as a base for lake colours in the dye and colour industry; as an accelerator and stabiliser of emulsions of various types, and in the preparation of asphalt, coal tar residues and pitch emulsions; as an absorbent of nitro-glycerine in the manufacture of dynamite; as a filler material for fertilisers; and for paints, rubber, and moulded composition products of various kinds; as a suspending agent in core washes for foundry work, and as a bonding agent in moulding sands; as a sticking or spreading agent in insecticidal sprays and dusts, and to increase the wetting power of cattle dips; as an ingredient of lubricants; in cold water paints or distempers; as an ingredient of adhesive pastes; as a sizing agent for textiles, cordage, etc.,; as an ingredient of crayons, indelible pencil leads and pastel colours, and of printers' inks; for increasing the retention of china clay in the manufacture of paper, and for removing the carbon black particles in the de-inking of newsprint; to remove the impurities from petroleum, gasoline and lubricating oils, as well as from packing-house products and vegetable oils (acid-treated clay); to replace whiting in putty; as a source of pure colloidal silica, or silica gel; as a detergent in

soaps and scouring compounds; as an ingredient in stove polish; in pharmaceutical products, cosmetics, facial clays, etc.

The above list by no means exhausts the many suggested uses for bentonite. Thus far, important commercial utilisation has been confined to the paper and oil industries, the latter being by far the largest consumer. The clay in this case, while probably more or less identical with bentonite, is variously termed halloysite or otalite, and is found in southern California: large quantities have been used by the Pacific Coast oil refineries, the clay being first treated with acid.

A Note on the Conduction of Heat Down the Necks of Metal Vacuum Vessels containing Liquid Oxygen, by G. R. D. HOGG, B.A. (Oxon.).

Griffiths has described the metal vacuum vessel used for the storage of liquid oxygen, and has discussed briefly the various causes to which thermal leakage into the interior. of such vessels may be attributed. This question has also been dealt with in the Report of the Oxygen Research Committee. These discussions show that the heat which enters cold liquids stored in vacuum vessels is transferred from the outer shell by three principal agencies, conduction through the residual gas in the vacuum space, radiation across the vacuum space and conduction down the walls of the inner neck which usually forms the only solid connection between the inner and outer vessels. It is possible by methods indicated by Griffiths to calculate approximately the amount of heat which will enter by the first two agencies, but the amount traceable to the

third agency is difficult to estimate, and there are considerable differences of opinion as to its magnitude.

In neither of the sources quoted is any attempt made to calculate the heat entry down the neck, and experimental data available are meagre and their accuracy doubtful. Both accurate calculation and experiment are rendered difficult by the fact that a stream of cold gas from the interior of the vessel and by abstracting heat from the inner walls of the neck of the vessel materially affects the temperature gradient, and therefore the flow of heat, in the neck. The magnitude of this stream of gas, and hence its cooling effect are, of course, directly proportionate to the total amount of heat entering the vessel, by whatever means, in unit time, and therefore depend not merely upon the size of the neck but also upon the other dimensions of the vessel,

Approximate methods are now given for calculating the "neck loss" of metal Dewar flasks containing liquid oxygen. The methods may also be used for determining the most suitable dimensions for the inner neck of a metal Dewar flask, the other dimensions of which are known.

The Mechanism of Setting of Calcium Sulphate Cements, by C. L. HADDON, A.I.C., A.INST. P., M.Sc.

Though the nature of the chemical processes in the setting of calcium sulphate cements has been known since the time of Lavoisier, no completely satisfactory explanation of the physical processes involved has yet been put forward, the difficulty being that each authority has endeavoured to explain the hydration of every variety of calcium sulphate to dihydrate by one all-embracing hypothesis, and in every case, facts not explicable by the particular hypothesis have prevented any explanation being generally accepted. The object of this paper is to put forward an explanation which covers the known facts.

Calcium sulphate cements are chiefly manufactured from gypsum. On heating under manufacturing conditions, gypsum loses three-fourths of its combined water at about 130° C., with the appearance of "boiling." At 160-170° C. the remaining water is almost all lost, and anhydrous calcium sulphate with a little hemihydrate is left. This product is "soluble anhydrite." More intense and prolonged heating produces more perfect dehydration. Above

1000° C. decomposition of the calcium sulphate occurs.

If

The stable form of calcium sulphate in contact with water at ordinary temperatures is the dihydrate CaSO,2H2O, and both the hemihydrate and anhydrous salt, on mixing with water, dissolve and reprecipitate as the less soluble dihydrate. the rate of hydration is sufficiently rapid, a paste of burnt gypsum and water sets to form an adherent mass, and the use of such cements has been known for centuries. This absorption of water occurs also on exposure of anhydrous calcium sulphate to the atmosphere, hemihydrate being formed. With very fine particles such as in "soluble anhydrite," a few weeks in damp weather is sufficient for complete conversion into hemihydrate.

Plaster of Paris is almost entirely hemihydrate. With water, it sets to a hard mass in ten to twenty minutes, and its rapid setting prevents its being mixed with water by means of a trowel. It is usually stirred into a thick paste with water, and the thick cream produced poured into a mould. The plaster (sp. gr. 2.7) is generally mixed with about half its weight of water. The density of dihydrate is 2.31, thus if neither expansion nor contraction occurred in the apparent volume of the mass, a mixture of two parts by weight of plaster with one of water would produce 41 per cent. of voids in the final dry mass, and the reaction,

2CaSO.H2O + 3H2O = 2CaSO,.2H,0,

results in a decrease in true volume of 7.7 per cent. Actually there is a very appreciable increase in apparent volume, a property on which the commercial use of plaster for making casts depends. Van't Hoff, by means of a pyknometer, showed there was actually in setting a steady decrease in true volume, followed by a slight expansion. Davis considers this to be due to the transition of an unstable ẞ dihydrate to the stable a or monoclinic form.

The completely anhydrous salt is found a sthe mineral anhydrite. On mixing with water this hydrates very slowly, and the dihydrate formed is not adherent.

Intermediate in setting properties between these two extremes is gypsum heated to 500° C. and finely ground. Its water content varies, but 0.5 per cent. is a conmon figure. Its setting speed depends on the fineness of grinding and the amount of

hemihydrate present, the time being of the order of six hours to acquire a tensile strength of 80 lb. per square inch. This speed is sufficiently slow to enable the plaster to be mixed with water by means of a trowel, and thus the proportion of plaster to be mixed with wthmarf dwolynup hmsc to water can be made considerably greater than with plaster of Paris, resulting in the production of a much stronger cement. It expands slightly on setting, but much less than plaster of Paris.

Le Chatelier claimed that on setting, the deposited dihydrate was in the form of spherulitic crystals, and the interlocking of these conferred on the mass its mechanical strength. Owing to the smallness of the crystals, he was unable to prove this in aqueous solutions, but in solutions of the hemihydrate to which alcohol had also been added and in which calcium sulphate is less soluble than in water, he obtained proof of the formation of such crystals. Von Weimarn showed that the nature of a precipitate depends on the solubility of the precipitate and the degree of supersaturation at the moment of formation. Great supersaturation yields gelatinous precipitates which subsequently crystallise. This may be classified as Type I. With medium supersaturation, spherulitic crystals are obtained (Type II.), while slight supersaturation results in the formation of welldefined single crystals (Type III.) The less the solubility, the greater is the tendency to form a gelatinous precipitate, and the smaller the supersaturation, the longer the time for complete precipitation and the larger the crystals formed.

There seems little doubt that the process of setting consists in the first place of the solution of plaster followed by its deposition as less soluble dihydrate. Ostwald and Wolski, and later Neugebauer, examining plasters containing much. hemi-hydrate, found that on mixing with water there was temporarily a large increase in viscosity.

The true explanation appears to be that of Desch, who instanced the practical use of sodium thiosulphate solutions in imitating the action of frost on rocks. Here the growth of crystals results in a very great disintegrating force corresponding to the action of water on freezing, though in this case the volume of the solid deposited is less than that of the original liquid. apparent expansion is due to the crystal growing in its true crystallographic form

The

rather than that of the cavity it occupies. It has been shown that set plaster is twice as strong wet as when dry, and the expansion of plaster kept thoroughly wet is much greater than when the mixed cement is allowed to set in a warm dry room and occasionally moistened. Thus the weaker the cement during setting the greater is the expansion. This result would be explicable if the expansion was due to crystal thrust.

The Vapour Pressure of Tellurium, by J. J. DOOLAN and J. R. PARTINGTON.

The vapour pressure of sulphur has been determined in various ways, and at least one determination of that of selenium has appeared. The vapour pressure of tellurium, however, has, apparently, not previously been measured, and an attempt has been made to determine its value at least approximately.

The method employed was essentially that used by von Wartenberg in his investigation of the vapour pressures of metals.

In this a measured quantity of inert gas is passed in a steady stream over the heated metal. The loss in weight of the metal was obtained at three fixed temperatures, at least three rates of flo wof gas being used at each temperature. From the relation

the element was obtained for each rate of flow, and by plotting these the pressure for zero rate of gas streaming was obtained for each temperature.

The quantity of gas used was determined by measuring the rate of flow and the duration of an experiment.

Nitrogen was chosen in preference to hydrogen as the inert gas because, although tellurium nitride has been prepared, it cannot exist above 200° C. On the other hand, if hydrogen were used, it is quite likely that traces of hydrogen telluride would be formed and vitiate the results.

Its flow was controlled with great accuracy by means of a venturimeter. The quantity of gas passing in each experiment was, therefore, exactly determined by the rate of flow and the time of duration of the periment.

A silica boat contained in a silica tube was used to hold the tellurium, and back

diffusion of the vapour was prevented by causing the gas to force its way through a tightly fitting plug, of boked fireclay, or, later, of uralite which had been previously maintained at a bright red heat.

Instead of having the thermocouple outside the tube, as in von Wartenberg's work, it was inserted in a silica sheath which fitted tightly into another resistant plug in the silica tube. In this way, during the passage of gas, the silica boat was enclosed in a tiny chamber in the middle of the heating zone of the electric furnace. By using a long silica tube the great advantage was obtained of having its ends quite cool. Thus, by attaching the inner plug to the weighed boat by a hook of nichrome wire and by affixing a second and similar plug to the thermocouple sheath, it was possible to draw the hot boat into a cooler part of the tube at the conclusion of an experiment and so allow it to cool in an atmosphere of nitrogen, without waiting for the whole furnace to cool before being able to withdraw and weigh the boat.

The coolness of the end also made luting and unluting a simple matter, so that the rate of flow of gas was readily checked, and the possibility of the development of an undetected leak obviated. Between experiments a current of gas was maintained for some minutes in order to remove any tellurium which had condensed on the plug or elsewhere in the heated region. The neces sity for this procedure was pointed out by von Wartenberg, who found that failure to do so gave results which became lower owing to the now partly saturated gas taking up less tellurium from the weighed boat. The results were given in a table.

Cryoscopy in Sodium Sulphate Decahydrate, by E. E. TURNER and W. H. PATTER

SON.

Cryoscopic and conductivity measurements have shown that the nitroprussides are derived from the acid H2Fe(CN), NO, and not H,Fe(CN)1,(NO)2: Burrows and co-workers have further shown that ferrocyanides and complex oxalates also possess simple formulae. The molecular weights of a number of sodium salts have now been determined, using sodium sulphate as cryoscopic solvent. The results of Löwenherz have, in general, been confirmed, and his experimental method has been adopted with only slight modification.

The depression of the transition point of

sodium sulphate produced by carbamide was found to be not strictly proportional to the concentration, a fact not recorded by Löwenherz. The deviation from proportionality was, however, insufficiently large to affect conclusions drawn in connection with the molecular complexity of solutes in general, since comparative, not absolute, molecular weights were required.

For K, the depression constant, the value 32.5 has been taken, partly from the results of the present work and partly from the results of previous workers.

The general reliability of the method was tested by determining the molecular weights of a number of organic compounds and of simple sodium salts. It was found necessary, in order to obtain sharp thermometric readings, to cause rapid initial crystallisation of the dodecahydrate from mixtures. this being effected by using a water-cooled air-jacket. Poor results were obtained when the jacket was kept at about 30°. In no case was the mixture allowed to supercool more than 0.3°.

Abnormal results were obtained for the molecular weights of sodium oxalate and borax. The low result for the former substance has been shown not to be due to solubility effects, whilst the molecular weight obtained for borax is in good agreement with the value obtained by Boutaric, Chauvenet, and Nabot.

The molecular weights obtained for sodium nitroprusside, ferrocyanide, aluminioxalate, ferrioxalate, and chromoxalate agree with those obtained by the measurements in aqueous solution referred to above.

Determinations made with sodium salts of d-tartaric and racemic acids show that racemic compounds do not exist in solution under the given conditions, although the latter would appear to be favourable for the persistence of racemic compounds derived from sodium.

The Catalytic Decomposition of Hydrogen Peroxide Solution by Animal Charcoal: The Production of Highly Active Charcoals, by J. B. FIRTH, D.Sc., F.I.C., and F. S. WATSON, M.Sc.

Ordinary blood charcoal which has been previously heated to 120° C. shows only moderate catalytic activity in the decomposition of hydrogen peroxide solution. The catalytic activity is considerably increased by previous heating in a vacuum at 600°

and 900° C., and is still further increased by previous sorption of iodine and subsequent complete removal of the iodine. It is shown that the activity of an activated charcoal consists of two types, one of which is very rapid and is termed a activity, and only exists for a few minutes, and a second type of activity, termed ẞ activity, which may persist for several hours. In ordinary blood charcoal which has not been subjected to activation treatment, the a activity is absent. It is also shown that the introduction of iron into sugar charcoal considerably increases its activity.

In the present communication the behaviour of animal charcoal of various degrees of purity, and various degrees of activity has been studied. One of the objects which the authors had in view in this series of experiments was to obtain the maximum degree of activity by the methods already described.

The animal charcoal used in the first series of experiments was an unpurified commercial variety as supplied by Merck. The charcoal was repeatedly digested with boiling water until all water soluble material had been extracted. The resulting product, on ignition, gave 83.6 per cent. of ash. The ash consisted mainly of calcium phosphate, and a small amount of iron. The resulting charcoal was then very finely powdered and treated by one of the following methods:

I. The finely divided charcoal was thoroughly dried in an air oven at 120° C.

II. The finely divided charcoal was introduced into a quartz flask, and heated in a vacuum for two hours at 600° C., allowed to cool in a vacuum, and the required amount of the charcoal rapidly weighed out.

III. As in II., except that the temperature of activation was 900° C.

IV. A quantity of the charcoal activated as in II. was treated with N/10 iodine solution in chloroform, in the proportion of 25 cc. of the solution per gram of charcoal, for twenty-four hours. The charcoal was then filtered off, transferred to a silica dish, and gently heated until practically the whole of the iodine had been volatilised. The charcoal was then shaken with an alcoholic solution of potassium hydroxide, and then boiled with distilled water until, on filtering, the filtrate showed no opalescence with silver nitrate solution. The resulting char