In the normal state of the atoms, two of the outermost electrons in ionized carbon appear to move in 2, orbits, while the clectron which generates the spectrum traverses a 2, orbit, as suggested by Bohr for neutral boron.

PROF. W. L. BRAGG, F.R.S. The Refractive Indices of Calcite and Aragonite.

Calcite and aragonite, two crystalline forms of calcium carbonate, are closely related in their optical properties. Both crystals possess strong negative birefringence, and their refractivities are nearly identical when allowance is made for the small difference in molecular volume between the two. The crystals have been analysed by X-ray methods.

The strong double refraction is to be ascribed to the peculiar form of the CO, group. This consists of three oxygen atoms grouped around a central carbon atom and lying in a plane. The oxygen atoms are more highly polarized by an electric field than the other atoms in the crystal. The three atoms in the same CO, group, when they become electrical doublets, exert a strong influence on each other, which results in their being more strongly polarized by an electric field parallel to the plane of the group than by a field perpendicular to the plane of the group. Hence the refractive index of the crystal is greater when the electric vector is perpendicular to the trigonal axis of calcite than when it is parallel to the axis. The same holds for the pseudo-hexagonal axis of aragonite.

A numerical calculation of the effect shows that it accounts quantitatively for the difference in refractive indices, both for calcite and aragonite. Values for the refractivities of the calcium and oxygen atoms are assumed which are based on the refractivity of a number of compounds. When these are substituted in the formulæ which are derived, the calculated values of the refractive indices do not differ by more than 1 or 2 per cent. from the observed values. Thus :·

hydride, and Succinimide. Communicated by Sir William Bragg, F.R.S.

The monoclinic holohedral crystal of succinic acid is shown to possess two inolecules per unit cell, and is based on the Bravais lattice Im Each of these molecules has a plane of symmetry, and they may be brought into coincidence with each other by a screw of angle 180° and translation 6: thus the space-group to which the crystal belongs is Ch2. A probable structure is discussed in detail, an explanation of cleavage and twinning being given.

The orthorhombic crystal of succinic anhydride is based on the lattice I', and has four molecules per cell. It is shown that the X-ray method fails to determine the class to which the crystal belongs; the space-group is either C', or Q'h, and the molecule is either asymmetric or possesses a plane of symmetry.

In the unit cell of the orthorhombic holohedral crystal of succinimide there are eight asymmetric molecules, the fundamental lattice being I. The space-group is proved to be Q'h, and details of a probable structure are given; a comparison of the molecules of succinic acid and of succinimide is made.

L. F. BATES and J. S. ROGERS. Particles of Long Range from Polonium. Communicated by Prof. Sir Ernest Rutherford, F.R.S.

Polonium has been examined by the scintillation method for the emission of longrange particles, and it has been found that for every 10 a-rays of range 3.93 cm. emitted there are also present 98, 51, and 26 particles of ranges 6.1 +0.1, 10+ 0.1, and 13.10.2 cm. respectively, together with about 7 particles of longer range, which are probably H-particles. It is considered from the relative brightness of the scintillations alone that these particles are a-rays. Also from the evidence available it is deduced that these particles are emitted by polonium itself.

A. MULLER. On the Determination of the Crystal-Axes in "Single-Crystal" Aluminium Bars by means of X-Rays. Communicated by Prof. Sir Ernest Rutherford, F.R.S.

In connection with the work of Mr. G. I. Taylor, F.R.S., and Miss C. F. Elam on the distortion of aluminium crystals during a tensile test (Bakerian Lecture, 1923), the

author investigated by means of X-rays the aluminium bars used by Taylor and Elam. It was found that these kept the attice structure even when stretched to a considerable extent. A description is given of the method developed for the determination of the orientation of the axes of the lattice relative to a fixed plane in the test-piece.

At the meeting of May 15, PROF. A. FOWLER, F.R.S., Yarrow Research Professor, will deliver the Bakerian Lecture on Spectra of Silicon at successive Stages of Ionization.

ne

analysis model is produced by making the height of each constituent proportional to its calorific value.

The production of inert constituents is fairly constant up to the seventh hour; carbon monoxide is evolved fairly evenly until the eighth hour, when it tapers off; while hydrogen gradually increases during the carbonising period until a maximum is attained about the fourth hour. Methane production is at its height in the first ten minues, and tapers off sharply until at the end of the second hour a fairly constant quantity is being produced. The unsaturated hydrocarbons, important though they are as regards their contribution to calorific value, are quite small in volume, and tail off at the sixth hour. Whereas the analysis curve shows the hourly production of each constituent of the gas, the analysis model demonstrates the contribution of each to the total thermal yield and also to the effective work at any one moment during carbonisation. It thus shows in a convincing manner the true value of each constituent. In the case of the inert constituents, nitrogen, carbon dioxide, and oxygen, it is obvious that as these possess no calorific value they will be represented without height in the model. The outstanding feature of this model is the layer representing the unsaturated hydrocarbons which disappear at the end of the sixth hour. This layer has

VOLUME & CALORIFIC VALUE CURVES

SHOWING EFFECT OF

HOLE IN RETORT CONTROL IN FULL LINE

Fig. 8.

shows the structure of such a model, and is composed of sections representing the different constituents of the gas. The height of each section represents the calorific value of the constituent, and the breadth, the volume produced at any one moment during carbonisation. By looking down on the model in plan, a simple reproduction of the stratified analysis curve is seen, and the

considerable height owing to the high calorifie value of the unsaturated compounds. It is practically a knife edge owing to their small volume.

Reverting to the soundness of the retort and its effect upon the nature of the gas made and the yield of gaseous therms, fig. 9 gives the results when working under standard conditions, but with a hole of in. diameter in the retort at a point equivalent to the centre of a 20ft. retort. The broken curves show (with the control curves) the changes in volume and calorific value resulting from the hole. Whereas the volume peak is reduced at the commencement, due to loss of rich gases, the volume is increased throughout practically the whole of the remaining period of carbonisation. This is probably due to more rapid carbonisation taking place as the result of the scouring action of the furnace gases pulled into the retort, and to the furnace gases themselves. On the other hand, the calorific value of the gas is lower throughout, that of the final gas being 482 B.Th. U., as against 532 B.Th. U. in the control experiment. But manufacture of gas does not stop at the tenth hour, for at this period there is being produced about 700 cubic feet per hour of a gas of 180 B.Th. U. per cubic ft. This is producer gas (see fig. 10).

is difficult to determine what is the precise thermal yield of true coal gas under these conditions, as it would be necessary to know at what stage the producer gas reaction commenced. Rough calculations from the analyses show that 1 to 2 therms of true coal gas have been replaced by producer gas therms.

The thermal models constituted a device for detecting any slight differences in the carbonising process, or for determining the effect of any adjustments of the conditions of working. For obtaining a high thermal yield it has been found necessary to reduce the vacuum in the hydraulic main of any one section to level gauge when some of the retort doors of that section are open for charging or discharging. As hourly charging is adopted, it is thus the practice to reduce to level gauge for 10 minutes during each hour. An examination of the thermal model will show that at these periods there is a definite indent in the volume curve, and that immediately following this, when the vacuum is re-established, there is a corresponding peak in the calorific-value curve. It is evident, therefore, that gas has been throttled-up during this period of level gauge. Generally speaking, this method of

The effect of the producer gas reaction is clearly shown. The large increase in nitrogen in the tailing-off gas is to be expected. but it is surprising that the time of contact in the retort is sufficient to convert the car

Fig. 11.

Again a quantity of rich gas is lost in the initial stages of carbonisation, while the volume of gas during the last few hours is increased. The calorific value of the gas is also lower than under normal conditions. The analysis model and curves, fig. 13, show that a very different set of reactions must have occurred. Air drawn in at the mouthpiece produces combustion of the gas

Fig. 14 shows how disastrous it is in a pierced retort to bring the vacuum to leve! when the retort doors are open for discharging and charging (fig.-15).

THE SOCIETY OF GLASS
TECHNOLOGY.

(Continued from page 175.)

Continuous High Temperature Measure ments in Glass Works, by W. M. CLARK, PH.B. (U.S.A.).

In the strictest sense of the term, there was probably no absolutely accurate method existing for constantly determining glasshouse temperatures. We could only approach this desirable aim by refinements of installation and careful maintenance, and this meant both equipment and maintenance expense. Therefore, in considering the subject of a pyrometer installation the glass manufacturer must balance the factors of what his demands for accuracy will be worth against the initial and operating costs of various systems. It is always advisable to instal both the indicating and recording type of instrument, side by side.

Experience at a number of modern well laid out plants showed that pyrometric measurements were reflected in the costs sheets. The morale of the furnace men was improved by eliminating the human errors

of judgment in guessing at temperatures and cost per unit of output decreased. With improvements in pyrometric equipment the tendency was for their more extensive use around glass works. The depreciation on such equipment was not excessive, about 10 per cent. per annum being a fair amount judging from successful installations, and as improvements came along they could be incorporated into the system, replacing features which became obsolete, and thus keeping the installation up-to-date. A well planned system would reduce overall costs in a year more than sufficient to pay for the investment, and often would leave several times this amount if used intelligently.

The most forceful argument for the adoption of pyrometer control in a glass works was on the score of economic savings, but the information thereby made available to the management and staff was also important in showing each man a tangible effect of his operation on the thermal conditions. The producer man saw that good gas was necessary to maintain temperature, the furnace man saw the result of proper draught regulation, the foreman of the finishing department could be satisfied whether the ware was being properly annealed, and the whole schedule of production was benefited.

Recent Advances in the Design of Temperature Measuring Instruments, Etc., by R. W. WHIPPLE, F.INST. P.

One difficulty in connection with disappearing filament pyrometers was to get two lamps which would be interchangeable. Methods for overcoming this difficulty were indicated. The speaker then said how greatly he had been impressed with what he had seen recently in the United States in the matter of automatic temperature control. He understood considerable work was being done in this connection in glass works, and referred to the works of Messrs. Keuffel & Esser, where lehr temperatures were controlled from 200° C. to 600° C. For every kind of glassware a special cooling curve was followed. In the United Kingdom automatic temperature control was being developed. The main principle in such control was a galvanometer needle, which was deflected by some method. Lantern slides were then exhibited, showing various forms of controller, including (1) Brown, (2) Bristol, (3) C. Engelhard. (4) Leeds &Northrup. (5) Cambridge, (6) Barr, (7) Althorpe. It should be possible to control lehr temperatures by some such auto