35* 



NA TURE 



[February 8, 1906 



the specimens were heated to 1000° C. and then chilled in 

 cold water, care being taken to support them during the 

 process, as cast-iron is very brittle at such a high tempera- 

 ture. The rods were tested for maximum remanent B and 

 coercivity by Madame Curie's method, the magnetised bar 

 being placed in a long solenoid producing a demagnetising 

 field which was gradually increased until a search-coil 

 slipped along the bar showed that the demagnetisation 

 was complete. The results showed the chilled cast-iron 

 to be not very inferior to ordinary magnet-steel. By 

 ballistic tests on the two rings, their permeability curves 

 were obtained, and these indicated that the simple process 

 of chilling used was quite satisfactory even for a tolerably 

 massive ring of 6 si] cm. cross section. The cheapness 

 and ease of working cast-iron should encourage instrument 

 makers to test its capabilities in various instruments. — 

 Experiments on the propagation of longitudinal waves of 

 magnetic flux along iron wires and rods : Prof. Lyle and 

 .Mr. Baldwin. The experiments described in the paper 

 were undertaken with the object of determining if there 

 is .1 definite rate of propagation of magnetism in iron. 

 The method adopted was to produce magnetisation at a 

 particular point on a bar by means of a coil through which 

 an alternating current was passed, and then to observe the 

 magnetic flux at various distances from the coil by means 

 of .1 small secondary coil, free I" be moved to various 

 places on the bar. By the use of Prof. Lyle's wave-tracer 

 the magnetic flux at various points along the bar was thus 

 obtainable. The wave curve was then analysed by 

 Fourier's series. Various curves given in the paper show 

 the value of tin- constants in Fourier'- series and of the 

 lag in the magnetisation as the coil was moved along the 

 bar. Contrary In what had been observed in previous 

 researches, tin- authors found that the phase lag, instead 

 of continuously increasing along the bar, reached a maxi- 

 mum value and then diminished, proving the absence of 

 true wave propagation. 



Mineralogical Society, January 23. — Prof. H. A. Miers, 

 F.R.S., president, in the chair. — Studies in crystallisation; 

 sodium nitrate: II. A. Miers and J. Chevalier. Micro- 

 scopic observations were made upon solutions of known 

 strength contained in open tubes or sealed tubes main- 

 tained at a known temperature, or in the form of drops 

 upon a slide, with the object of comparing the growth 

 of crystals in metastable and labile solutions respectively. 

 The limits of the labile state (in which the solution tan 

 crystallise spontaneously) have been fixed by previous ex- 

 periments by II. A. .Miers and Miss F. Isaac. II a crystal 

 of the salt lie introduced into a supersaturated solution 

 which I- not labile, the centres of growth of new crystals 

 are on its surface, and they grow in parallel positions 

 upon it ; if it be introduced into a labile solution the new 

 centres of growth are in its neighbourhood, and the crystals 

 fall upon it in various positions. If it be moved about in 

 either, a cloud of crystals is produced; but in the meta- 

 stable solution this appears to be due to minute crystals 

 which are swept from its surface. A crystal having 

 appeared spontaneously can continue to grow in a labile 

 solution without producing others in its neighbourhood, 

 but if introduced it at once produces a cloud. This may be 

 because the growing crystal is surrounded by a zone of 

 metastable solution. — Geikielite and the ferro-magnesian 

 titanates : T. Crook and B. M. Jones. Geikielite occurs 

 in association with magnesian menaccanite and common 

 ilmenite (menaccanite) in tin- gem gravels of the Balan- 

 goda and Rakwana districts of Ceylon. A considerable 

 number of analyses indicate that geikielite varies in com- 

 position, the iron oxides ranging from 8 per cent, to 14 pet- 

 cent. No specimen has hitherto been found which contains 

 less than 8-1 per cent, of iron oxide. For this reason the 

 formula (Mg,Fe)TiO., is preferable to MgTiO, as express- 

 ing the true composition of geikielite. Magnesian 

 menaccanite containing about 28 per cent, of iron oxide 

 i, very 1 losely allied to geikielite in all its properties, more 

 so than to common ilmenite. The alteration products of 

 geikielite are similar to those of ilmenite, consisting of 

 rutile and so-called leucoxene ; the latter is a mixture of 

 amorphous titanic acid, sphene, and limonite. It seems 

 advisable to classify the ferro-magnesian titanates as 

 ilmenites and geikielites, treating magnesian menaccanite 



(which has the formula (Fe,Mg)Ti0 3 , where Fe : Mg = i : 1) 

 as the middle member of the series. — G. F. Herbert Smith 

 exhibited, and explained the use of, a diagram for the 

 graphical determination of the refractive index from the 

 prism angle and the angle of minimum deviation. He also 

 explained a simple test for ascertaining the pair of faces 

 corresponding to any refracted image. 



Dublin. 

 Royal Dublin Society, December 19, 1905. — Dr. R. F. 

 Scharff in the chair. — The causes of " blowing " in tins 

 of condensed milk : Dr. G. H. Pethybridge. Blowing 

 (i.e. bulging) is caused by the accumulation of gas pro- 

 duced by the fermentation of the cane sugar added during 

 manufacture by certain wild yeasts or torulae, which can 

 ferment saturated solutions of sugar, and appear to be 

 present in the original milk supplies, and are not intro- 

 duced during the process of manufacture. — Two new 

 species of Collembola for Ireland : Prof. G. H. Carpenter. 

 The species described belong to the genera Isotoma and 

 Entomobrya, the latter showing some interesting special 

 affinities with Orchesella. 



Edinburgh. 



Royal Society, January 8. — Prof. Crum Brown, vice- 

 president, in the chair.- Bathydraco Scotiae, Poisson 

 abyssal nouveau recueilli par l'Expedition Antarctique 

 national Ecossaise. Note preliminaire : Louis Dollo 

 The genus Bathydraco was instituted by Gunther in 1878 

 for a small fish (B. antarcticus, Giinth.) from the south 

 east of Heard Island, inhabiting a depth of 1260 fathoms. 

 The new species, named by M. Dollo, was obtained in the 

 Weddell Sea at a depth of 1410 fathoms. — Influence of 

 thymus feeding on allantoin excretion : Dr. J. Mac- 

 lachlan. The work was based on experiments carried 

 out in the laboratory of the Royal College of Physicians 

 of Edinburgh. Reference was made to the very unsatis- 

 factory nature of the evidence on the influence of uric 

 acid and nucleins on the production of allantoin, and it 

 was pointed out that the administration of thymus sub- 

 stance was invariably followed by a large production of 

 allantoin. The point investigated was whether this was 

 due to the conversion of the nucleins and puric bodies 

 contained in the thymus or to some specific action of the 

 substance. Boiling the thymus before it was administered 

 reduced its power of producing allantoin to less than 

 half. Thymus also was found to exercise a much more 

 marked effect in causing the production of allantoin than 

 other glands, such as pancreas, liver, and lymphatic 

 glands, which are also rich in nucleins. The conclusion 

 drawn was that raw thymus when administered produced 

 a specific action on the metabolism by which the formation 

 of allantoin was increased. — A theorem in hyper-complex 

 numbers : J. H. Maclagan Wedderburn. A short proof 

 was given of a theorem, first proved by Scheffers, that 

 an algebra which contains the quaternion algebra can, if 

 the moduli of the two algebras are the same, be expressed 

 as the product of the quaternion algebra and another 

 algebra. The theorem was then extended to a large and 

 important class of algebras. 



January 22. — Lord M'Laren, vice-president, in the chair. 



A form of initiational disturbance more convenient than 

 that of §§ 3-31 of previous papers on waves: Lord 

 Kelvin. The investigations of §§5-31, including the 

 " front and rear" of infinitely long free processions of 

 in deep water, are all founded on superposition of 

 equidistant initiational disturbances, the first of two 

 typical forms described in §§ 3, 4. In this form the 

 initial disturbance is everywhere elevation or everywhere 

 depression, and its amount at great distances from the 

 middle varies inversely as the square root of the distance 

 from a horizontal line at a small height above the water 

 surface in the middle of the disturbance. A type-disturb- 

 ance derived mathematically from that used in §§ 5-31 

 In double differentiation with reference to time, or by 

 single differentiation of the second of the forms of §§ 3, 4 

 with reference to space, is given in the present paper, and 

 illustrated by diagrams of curves placed before the society, 

 in which the initial disturbance has as much water above 

 as below the undisturbed level : and at great distances the 

 depression or elevation varies inversely as the 3/2 power 



NO. T893, V ° T - 73] 



