May 12, 1921] 



NATURE 



345 



segregate can be cast in a satisfactory manner pro- 

 vided that a suitable temperature is employed, that the 

 mould is properly designed and made, and that the 

 metal is skimmed and poured in the right manner. 

 Pure metals or single substances, as a general rule, 

 possess the sa^e density whether slowly or quickly 

 cooled. Those alloys in which there is an interval 

 between the liquidus and solidus solidify over a tem- 

 perature range, and often expand when slowly cooled, 

 as, for instance, when cast in sand. Extenspmeter 

 tests by Prof. Turner have shown the nature and 

 extent of such expansions in a number of typical 

 alloys. These results agree with the density deter- 

 minations, but the extensometer has the advantage of 

 showing the sequence and amount of each volume- 

 change. In practice one of the chief causes of failure 

 is pouring at too high or too low a temperature. 

 Other causes include such troubles as imperfect or 

 badly fixed cores, faults in moulding, cracks, mis- 

 runs or run-outs, and breakage in handling. In 

 foundry work generally the losses from all causes 

 reach about 10 per cent, of the output. Prof. Turner 

 took the view that casting is fundamentally an art, 

 and the part of the man of science will be to intro- 

 duce new ideas and processes rather than to improve 

 on present technique. 



The Oxford University Press is to publish under 

 the title of " From a Modern University : Some Aims 

 and Aspirations of Science " a volume of occasional 

 addresses by Prof. A. Smithells. 



A LENGTHY Catalogue (No. 88) of botanical and horti- 

 cultural works which are for sale by Messrs. Dulau 

 and Co., Ltd., 34 Margaret Street, W.i, has just 

 reached us. It gives particulars of no fewer than 

 3017 publications, including the library of the late 

 M. Edouard Andre, of Paris. Many early printed 

 and rare herbals are listed, and practically the whole 

 range of botany and horticulture is represented. The 

 catalogue should be seen by all who wish to augment 

 their libraries in these branches of science. 



Messrs. G. Bell and Sons are shortly adding a 

 new volume to the advanced section of their mathe- 

 matical series, namely, "A First Course in Statistics," 

 by D, Caradog Jones. The fundamental importance 

 of the right use of statistics is becoming increasingly 

 evident on all sides of life, social and commercial, 

 political and economic. It is hoped that a study of 

 this book will enable the reader to discriminate be- 

 tween the masses of valuable and worthless figures 

 published, and to use what is of value intelligently. 



Our Astronomical Column. 



Eclipse of Rhea by Titan. — The computing section 

 of the British Astronomical Association, recently 

 formed under the directorship of Mr. L. J. Comrie, 

 undertook an extensive examination of the phenomena 

 of Saturn's ring and satellites about the time when 

 their planes are turned edgewise. In the course of 

 this work the prediction was made that the very rare 

 phenomenon of the eclipse of Rhea by the shadow of 

 Titan would take place on April 8. A number of 

 members of the association, including the president, 

 Major Hepburn, observed the phenomenon, and found 

 a satisfactory accord with the prediction. Rhea faded 

 rapidly at loh. 22m., and became invisible for 50 

 minutes ; the estimated time of mid-eclipse was 

 loh. 47-3m. G.M.T. ; the corresponding predicted 

 time was loh. 38m., and predicted duration 44 minutes. 

 Since the relative motion of the satellites was slow the 

 error in their positions is small. 



It would be well if the national ephemerides could 

 publish predictions of interesting phenomena of this 

 character, as there is a danger of their escaping 

 notice if left to unofficial agencies. There is probably 

 no other observation of the kind on record in the 

 Saturnian system; even in the Jovian system, where 

 mutual eclipses occur more often, ver\- few have been 

 recorded, obviously because they have never been 

 systematically predicted. 



Comets. — L'Astronomie for April contains a re- 

 discussion by Mr. G. Neujmin of the orbit of the 

 comet (1916a) discovered by him in 1916. He finds : — 



T = 1916 March 1 1 -3239 G.M.T. 

 «=I93' 47' 33" 

 ^ = 327° 33' o" 

 / = 10° 37' 12" 

 <t> = 34° 26' 33" 

 fi =655-364" 

 Period 5 4 14 years. 

 The observations used extended from February 27 to 

 June 5. The comet should be in perihelion again 



NO. 2689, VOL. 107] 



about August 10 of the present year, but so unfavour- 

 ably placed that it is to be feared that it will escape 

 detection. 



Comet Pons-Winnecke has been deviating from the 

 predicted path with unexpected rapidity, and M. Ebell 

 has deduced the following revised orbit from observa- 

 tions on April 12, 16, and 26 : — 



T = i92i June i3'o8 

 <o =171° 437' 



Q= 96^ 38-9' 



z=i8° 15-0' 

 log ^ = 0*0152 

 log rt = o-4526 



The value of log a is almost certainly much below 

 the truth, but the elements will probably represent the 

 motion for the next few weeks. The ephemeris 

 printed in this column needs to be corrected by — 13m., 

 -1° 11' on May 14, and -25m., -1° 12' on May 20. 

 It will be noticed that the perihelion point is now 

 placed well outside the earth's orbit, which makes 

 the occurrence of a meteor shower somewhat doubtful. 



Mav 



June 



1916*0 



