126 



NATURE 



[April i8, 191 8 



period. Mr. Jeffreys will surely admit that a periodic 

 disturbance of this kind, acting parallel to the minor 

 axis of the orbit, would certainly affect the longitude 

 of perihelion, without atlecting the eccentricity ; 

 thoug-h whether the amount of resistance to be expected, 

 say from matter in the Zodiacal light, is sufficient to 

 make the effect appreciable may well be doubted. 

 Moreover, I had not thought of the resisting medium as 

 revolving in a planetary manner. I am inclined to 

 attribute mych more importance, to my other sugges- 

 tion based on the electrical theory of matter {Phil, 

 ^ag. for August, 19 17). Nevertheless, a periodic resist- 

 ance hypothesis is peculiarly applicable to Mercury, 

 (a) because of its nearness, (b) because of the eccen- 

 tricitv of its orbit. Oliver Lodge. 



Relativity and Gravitation. 



A MATHEMATICAL friend with whom I have been dis- 

 cussing Prof. Eddington's paper on "Relativity and 

 Gravitation," recently published in Nature, has made 

 what appears to me to be an interesting suggestion. 

 Prof. Eddington states that if a current of aether were 

 moving vertically (say) with a velocity of 161,000 

 m./sec, a rod 8 ft. long, when placed transversely to 

 the stream {i.e. horizontally), would, when turned ver- 

 tically, be only 4 ft. He also says that this contrac- 

 tion would be. unobservable because the retina of the 

 eye would have similarly contracted in a vertical 

 direction. Suppose, however, that the rod in its two 

 positions were observed, not directly, but by means of 

 a mirror inclined at an angle of 45°, by a spectator 

 lying on his back on the fioojr of the room ? His retina, 

 being horizontal, would, ex hypothesi, have under- 

 gone no contraction at all. Both images of the rod, 

 in its horizontal and vertical positions, would fall on 

 this horizontal retina. If the experiment could be per- 

 formed the contraction of the I'od ought to be evident, 

 and, afford direct proof of the Lorentz-Fitzgerald hypo- 

 thesis. Is there any flaw in this reasoning-? 



H. H. O'Farrell. 



It is interesting to examine Mr. O'Farrell's plan 

 for defeating the conspiracy to conceal the change of 

 length of the rod ; but the resourcefulness of the con- 

 spirators is equal to the occasion. A compensation 

 will take place in connection with the reflection of the 

 light from the moving mirror. Light rebounds from 

 a fixed mirror as though it were a billiard ball re- 

 bounding from a perfectly elastic cushion. If the 

 cushion were moving with a great velocity the angle 

 of rebound would naturally be modified. That is 

 only an analogy, but it will perhaps show that we 

 cannot apply the rules of elementary optics to the 

 formation of images by a mirror moving through the 

 aether. A mathematical discussion, on the basis of 

 Huygens's. principle, shows that a change of size 

 of the image will be introduced which compensates for 

 the change of size of the rod. It may be remarked 

 that in order to deflect the ray from the horizontal 

 to the vertical direction the mirror, although apparently 

 inclined at 45° to the horizontal, would actually (in 

 terms of the "real" space) be inclined at 266° 

 (tan-^ ^) ; this illustrates how the laws of reflection 

 become modified in the conditions postulated. 



A. S. Eddington. 



Elliptical Haloes. 



The accepted explanation of the haloes of 22° radius 

 which are seen surrounding the sun and moon implies 

 that they are exactly circular in form. About two 

 years ago, however, I noticed a halo which appeared 

 to be elliptical with the major axis vertical. I was 

 unfortunatelv unable to take any measurements on 

 NO. 2529, VOL. lOl] 



that occasion, but on March 18 last a lunar halo, which 

 iwas visible for a considerable time during the even- 

 ling, also appeared to possess a decided, though 

 slight, ellipticity. That this deviation from the circu- 

 lar form was not an illusion I was enabled to verify 

 by noting the positions of Capella and 7 Geminorum 

 ■relative to the ring. 



At 7.30 p.m. Capella appeared to be exactly upon the 

 inner edge of the halo, while y Geminorum was within 

 the ring at a distance from it, which, as nearly as I 

 could judge, was a quarter of the moon's diameter. 

 From these data I find that the radii of the halo 

 measured from the centroid of the illuminated disc of 

 the moon through these two stars were 22-8° and 21-4° 

 respectively. Assuming that the halo was elliptical with 

 ithe major axis vertical, I deduce values of 23-3° and 

 21-4° for the semi-major and semi-minor axes. I am 

 aware that a more or less complete halo the major axis of 

 which is horizontal is occasionally seen surrounding 

 the 22° halo, but records of haloes elongated vertically 

 are rare. In 1908 Prof. Schlesinger noticed one the 

 axes of which were about 7° and 4°. 



Sir Napier Shaw informs me that very little is done 

 in this country on the shapes of haloes, so that this 

 letter may serve to direct attention to the desirability of 

 obtaining accurate measurements. J. B. Dale. 



Craigness, New Maiden, Surrey, April 10. 



Abnormal Catkin of Hazel. 



In February last one of my students. Miss M. Ben- 

 son, brought me a flowering branch of the hazel 

 (Corylus Avellana) in which one of the male catkins 

 had a group of female flowers at the base. The other 

 catkins were entirely male, but this one had eight 

 female flowers, all of which appeared to be normal in 

 structure ; they were arranged like the male flowers 

 on the same axis, but the bracts had the pointed shape of 

 those of the ordinary female flower, and no other 

 peculiarity was perceived. It would seem that this is- 

 the adoption of the arrangement which is the normal 

 one in Castanea, but such cases appear to have been 

 rarely observed in this species of Corylus, although 

 known to occur in C. tubulosa. The bush was one 

 (growing on the bleak heathv moorland of Sutton 

 Coldfield. ' W. B. Grove. 



Birmingham. 



VOLCANIC STUDIES.^ 



THE death of Tempest Anderson in 1913, when 

 returning' from a voyag^e to the volcanoes 

 of the East, removed a very familiar fig-ure from 

 scientific circles. For years he had made a special 

 study of recent volcanoes, and as he was a very 

 hig-hly appreciated lecturer and delighted to ex- 

 pound his subject to popular audiences, there were 

 very few who take an interest in geology and 

 geography who had not had the pleasure of lis- 

 tening to him. He was a skilful and enthusiastic 

 photographer, and his lectures were illustrated 

 with beautiful lantern slides ; hence it is probably 

 correct to say that no one did more to inform the 

 public on the subject of volcanoes during the 

 twenty years before his death. At the Royal 

 Geographical Society, the Geological Society, the 

 British Association, and many local societies he 

 was always sure of a warm welcome ; and his 



1 " Volc.inic Studies in Many Lands." Being Reproductions of Photo- 

 graphs taken by the Author. By Dr. Tempest Anderson. The Text by 

 Prof. T. G. Bonney. Second Series. Pp. xv-i-88. (London : John Murray, 

 1917.) Price 15J. net. 



