October 26, 1905] 



NA TURE 



629 



the handsome coloured plates of blue ground and 

 diamonds of various shapes and colours, presented by 

 Mr. Gardner Williams, stand out conspicuously. 



While it is evident that much has been achieved, it 

 is equally certain that in some branches only a start 

 has been made. In fact, the dominant feeling pro- 

 duced by reading the several interesting articles is 

 one that should inspire the greatest hope and 

 enthusiasm among scientific students in this country 

 and throughout .South Africa. Here lie new worlds 

 of unknown possibilities. As vet we stand only on 

 the threshold. Far off glimpses of a wonderful 

 country have been obtained, but it is the sight of a 

 Kilimanjaro enshrouded in mist, not of the unclouded 

 mighty mouritain-mass. W. G. 



Stone Gardens. By Rose Haig Thomas. Pp. xii 

 and plates. (London : Simpkin, Marshall, Hamil- 

 ton, Kent and Co., Ltd., 1905.) 

 An old wall sheltering such plants as are accommo- 

 dating enough to grow in such a situation is often 

 a delight; but to undertake the formation of a " stone- 

 .gardcn " in the way suggested bv the author is to 

 run counter to all our notions of the amenity and 

 purpose of a garden. Various " designs " are offered 

 for adoption, such as a lyre-shaped outline made of 

 paving stones with flower-beds representing the 

 strings, and separated by narrow strips of stone. 



Another design shows three snakes intertwined, 

 each snake made of flat stones of a different tint from 

 its neighbour. The spaces between the serpentine 

 convolutions are filled in with flower-beds. Other 

 designs are more appropriate to a formal or archi- 

 tectural garden. 



Of course, there is no disputing upon points of 

 taste, and each garden-lover must exercise his or her 

 fancies according to circumstances and in obedience 

 to individual proclivity. But if the designer intends 

 to furnish a model for other people to adopt, then we 

 expect there will be comparatively few garden-lovers 

 who will share the author's taste or feel inclined to 

 adopt her suggestions. 



Be this as it may, the author gi\-es very clear direc- 

 tions as to how her designs should be carried out, and 

 very judicious instructions as to the plants to be 

 selected and the method of planting them. Provided 

 these be properly carried out, kindlv nature will do 

 her best to conceal the flags and stones, and if the 

 author's designs are somewhat interfered with in the 

 process, that will not be a matter for regret on the 

 part of most garden-lovers. The work is in quarto, 

 with fourteen designs in colour. 



Oblique and Isometric Projection. Bv John Watson. 



Pp. iv + 5q. (London : Edward Arnold, n.d.) 



Price 3.S. 6d. 

 In defining the forms and dimensions of solids by 

 means of scale drawings, a very useful method in 

 certain cases is that of metric projection whereby 

 three systems of parallel edges of the solid are re- 

 presented on paper by lines parallel to three axes 

 drawn in arbitrarily selected directions, and to any 

 three scales also independently chosen. The author 

 deals only with isometric projection, and considers two 

 cases, first, when the projection is orthogonal, 

 secondly, when the projectors are oblique with the 

 plane of projection taken parallel to a face of the 

 solid, so that figures parallel to this face appear 

 without distortion. The best part of the book is 

 probably the chapter giving examples, mostly of joints 

 in woodwork, used by the author in conducting 

 classes in manual training; but it is doubtful whether 

 it was worth while to publish a book of such limited 

 scope. 



NO. 1878, VOL. 72] 



LETTERS TO THE EDITOR. 



[Tlie Editor does not hold himself responsible for opinions 

 expressed by his correspondents. Neither can he undertake 

 to return, or to correspond with the writers of, rejected 

 manuscripts intended for this or any other part of Nature. 

 No notice is taken of anonymous communications. \ 



Eclipse Phenomena. 



No opportunity for discussion was given at the Royal 

 Society meeting last Thursday, but the following brief, 

 notes may be suggestive and possibly useful. 



The particles in the corona which reflect solar light to 

 us are presumably moving very fast away from the sun, 

 and accordingly are illuminated by light of apparently 

 extra-long wave-length. This light, thus lowered in re- 

 frangibility, they will emit ; and inasmuch as they are 

 probably moving at all sorts of speeds, we might expect 

 that Fraunhofer lines would be encroached upon and blotted 

 out from the resulting emission, especially as some particles 

 would have a component of velocity towards us and others 

 away from us. 



If any of the particles are emitted with anything like 

 the speed of some of those from radium, the maximum 

 change of frequency to be expected would be great. 



Particles illuminated by rays normal to our line of sight 

 will send us a plane polarised beam, but when the illumin- 

 ating rays are oblique to the line of sight, as may be the 

 case from some of the longer streamers, then the polar- 

 isation would be only partial. 



How far single electrons may be able to resist the forced 

 vibrations of light-waves, and thus become themselves 

 polarised sources, is a matter on which I hope to try some 

 experiments. The illumination in which they are immersed 

 near the sun is very intense. 



The circular or ring appearance seen in the midst of 

 the corona in some photographs, with geometric centre at 

 a distance from the apparent centre of explosion, looked 

 to me like a gigantic vortex ring. I see no reason why a 

 sun-spot should not eject such rings occasionally. 



Oliver Lodge. 



Geometry of Position. 



In connection with the review of Mr. Wilson's recent 

 book, on p. vi. of your supplement last week, may I direct 

 the friendly attention of the reviewer and your readers to 

 an old paper of mine in the Philosophical Magazine for 

 November, 1875, where some of the theorems referred to 

 are given. I myself have found a slight modification of 

 the rapid system of writing chemical formulae there 

 advocated, extremely useful, and should like to advocate 

 its use by elementary students of organic chemistry — but 

 that is another matter. Oliver Lodge. 



October 20. 



Eclipse Predictions. 



The discrepancies referred to by Mr. J. V. Buchanan 

 (p. 603) as existing between the French and British pre- 

 dictions for the recent total eclipse of the sun are due simply 

 to the fact that a different value of the moon's diameter 

 is adopted in the Connaissance des Temps from that in the 

 Nautical Almanac, the former being about 2"-7 greater than 

 the latter. Hence the breadth of the zone of totality and 

 the duration of totality on the central line are greater in 

 the French than they are in the British ephemeris. But 

 there is no occasion to impute mistake to the French 

 calculators. They merely assume a value of the moon's 

 diameter that is, in my opinion, too large for eclipse 

 purposes. A. M. W. Downing. 



October 20. 



Chelifers and House-flies. 



It may be that the view suggested in my letter to 

 N.4TURE of August 31, that the association of the Chelifcr 

 with the house-fly is to the advantage of the former in 

 providing it with a wider geographical distribution, is not 

 sound. I believe it is, but at the same time admit that 

 there is not sufficient evidence at present to prove that the 

 association is of material advantage to the species. 



The important point to determine, however, is whether 



