448 



NATURE 



; January i, 1920 



in Greek Lands," F. H. Marshall; and 'T.ife and 

 Labour in the 19th Century," C. R. Fay. yiessrs.. 

 Constable and Co., Ltd., announce " Montessori 

 Experiments," Miss Blackburn. Prof. Patrick- Geddes 

 has written a volume, which Messrs. Longmans and 

 Co. will issue shortly, on the life and work of Sir 

 Jagadis Chandra Bose, the founder of the Bose 

 Research Institute in Calcutta. 



OUR A.STRONOMICAL COLUMN. 

 Fireball on December 25.— A brilliant fireball was 

 visible on Christmas night at loh. 21m. at Bristol. 

 It must have very much exceeded Venus in lustre, for 

 it gave a flash which illumined the whole skv, and in 

 that section of its flight where the greatest' outburst 

 occurred it left a streak about 3° long for 40 seconds. 

 The apparent path was from ii5°+34° to io:;|°+i°. 

 The motion was rather swift, the course of about 35° 

 being traversed in 2 seconds. The radiant point is 

 doubtful: it mav have been at i65°+73°, 2i9° + 7r° 

 245°+72°, or 26i° + 62°. If the second "is the correct 

 position, the meteor may quite oossiblv be considered 

 to have been a fragment of Mechain-Tuttle's comet, 

 which has a period of about 13* vears. Further 

 observations of the object would 'be valuable, and 

 should be .sent to Mr. W. F. Denning, 44 Egerton 

 Road, Bristol. 



Comets.— The following continuation of the 

 <phemeris of Finlay's comet is for Greenwich "mid- 

 night, frtfm the elements in Lick Bull. 325 •— 



R.A. 



N. Decl. 



Log: 



Log a 



Jan. 2 ... 3 3 35 21 47 0-1703 9-8i6o 

 6 ... 3 12 57 22 24 o-i8o6 9-8510 



'4 

 18 



3 22 24 22 58 0-1911 9-8849 

 3 31 21 23 27 0-2013 9-9173 

 3 40 9 23 55 0-2114 9-9485 

 The comet will traverse the Pleiades on January 18. 



It is calculated that Holmes's comet passed peri- 

 ht-hon about November 30, and a search ephemeris 

 was published. The comet is nrobablv too faint to 

 give much hope of its recovery. It has' not been seen 

 for two revolutions. 



Radiation Pressure.— The Asirohhvsical Journal 

 for October last contains an article by Mr. Megh Nad 

 .Saha in which the opinion is expressed that the 

 quantum theory of light will explain' the repulsion of 

 particles much more minute than those the dimen- 

 sions of which are of the order of a wave-length In 

 the undulatory theory the repulsion is a maximum 

 for particles of that order of magnitude, and becomes 

 practically zero for those of the dimensions of mole- 

 cules. Mr. .Saha quotes the results of spectrum analysis 

 of comets' tails, and some laboratory experiments' by 

 Lebedew (Ann. der Phvsik, 1910), for the fact that 

 gaseous molecules actually do sufler repulsion by 

 r.idiation pressure, which he considers an argument in 

 favour of the quantum theory. 



.\ssuming that a pulse 'of light gives all its 

 momentum to a hydrogen atom, the velocity imparted 

 to the latter by each "kick" would be 60 cm. /sec. 

 .Some calculations are given, from which tlie author 

 deduces that by repeated "kicks" the atom might 

 acquire a velocity of 6x10' cm. /sec, which has some- 

 times been observed in the solar prominences. 



The Orion Nebula.— We lately noted Dr. Berg.. 



strand's estimate of the parallax of' this object, 0-0078". 



Prof. VV. H. Pickering (Pubns. .Ast. Soc. Pac. April. 



19 19) contends for the value 0-0020". This is deduced 



NO. 2618, VOL. 104] 



from assumptions of the absolute magnitudes of a 

 number of faint stars which appear to be associated 

 with the nebula. By comparing their photographic 

 with their visual magnitudes, he concludes that their 

 .spectral type is A or B, whence their absolute magni- 

 tude is unlikely to be very low. But this involves the 

 conclusion that the brighter stars in Orion are super- 

 giants. Rigel in particular would have 87,000 times 

 the luminosity of the sun. But perhaps it is nearly 

 as easy to accept this as the value 5000 times the sun, 

 which results from Dr. Bergstrand's parallax. Prof. 

 Pickering estimates for the masses of the faint B7 

 stars in the nebula only four times that of Jupiter, 

 using his parallax. With Kapteyn's parallax 0-0054", 

 the mass would be one-twentieth of this. Either value 

 seems far too' small for a body to attain the tempera- 

 ture necessary to shine as a B star. 



SPHERICAL SHELL CRYSTALS IN 

 ALLOYS. 



AT the autumn meeting of the Institute of Metals 

 recently held in Sheffield, Dr. J. E. Stead pre- 

 sented an account of his investigations on some 

 ternary alloys of tin, antimony, and arsenic, one of 

 which was noticed by him to crystallise in a most 

 unusual and remarkable way. 



Having found that the alloys of antimony and tin 

 crystallise in what appear to be cubic crystals, and those 

 of tin and arsenic in rhombohedral flat plates, he made 

 trials with the object of finding how the metals would 

 arrange themselves when the three elements were 

 fused together and the melt allowed to cool. The results 

 obtained were astonishing, for the crystals found in 

 the matrix had the form of incomplete spherical shells, 

 the radii of which were small or great, according to 

 the time allowed for development. With rapid freezing 

 the radii were less than half a millimetre; when it was 

 protracted for one hour they were 5 mm. or more. 

 The most perfect structural arrangement of th(; 

 crystals was obtained in an alloy containing from 

 70-85 per cent, of tin, 25-15 per cent, of antimony, 

 and 4-5 per cent, of arsenic. Whether cooled slowly 

 or quickly, the polished surface of the alloys, after dis- 

 solving away the matrix, is very suitable for printing' 

 blocks, since the hard crystals stand out in bold 

 relief (see Fig. i). The alloys are very brittle, and 

 the fracture was found to travel midway along the 

 shell walls. An alloy containing tin 70 per cent., 

 antimony 25 per cent., and arsenic 5 per cent, gave 

 the following arrests On cooling : — 



(i) First separation of crystals ... 440° C. 



(2) Retardation in cooling between 325° and 320° C. 



(3) Solidification of the eutectic ... 244-9" C. 



The last-named temperature agrees closely with that 

 of the eutectic of the tin-antimony alloys. The con- 

 clusion is, therefore, warranted that the eutectic can- 

 not contain more than a trace of arsenic, an inference 

 confirmed by experiment. It was afterwards shown 

 by analysis that the primary crystals contain a maxi- 

 mum dmourit of arsenic, and that, as crystallisation 

 proceeds,>ithe defKJsits contain less and less of this 

 metaL i s: - 



, A large number of ternary alloys were prepared. It 

 was found that, while it required 2-5 per cent, of 

 arsenic in the presence of 25 per cent, of antimony to 

 produce slightly curved crystals, 0-5 per cent, of arsenic 

 in the presence of 3-75 per cent, of antimony yielded 

 curved segments in the upper layers. In an alloy con- 

 taining 1-65 per cent, of arsenic, 14-35 P*'' cent, of 

 antimony, and 85 per cent, of tin, spherical crystals 

 were found in the top layers, below these smaller seg- 



gave 



