April 25, 1901] 



NA TURE 



607 



Pennant, i.e. the first part of the " Natural History of 

 Selborne," freely interpolated with bracketed addenda 

 from the originals in the British Museum, and including 

 three or four letters of which White did not make use in 

 preparing his book for the press. Then, pleasantly intro- 

 duced by Dean Hole, and occupying more than 200 pages, 

 come White's garden diaries from 1751 to 1771 ; of which 

 a specimen, and enough to give an idea of White's personal 

 activity as a gardener, was printed as an appendix to Bell's 

 edition of the "Selborne " in 1877. The second volume 

 contains the letters to Barrington, also with additions and 

 interpolations from the originals, except in the case of 

 the famous " monographies " of the Hirundinidae, which 

 were published separately by White in the Philosophical 

 Transactions ; the antiquities of Selborne are also here, 

 and at the end we find a bibliography and a useful index, 

 which appear to be sufficiently complete. Each volume 

 is profusely illustrated. Mr. Keulemanns' drawings of 

 birds are familiar and welcome ; we have also a large 

 number of fancy sketches by Mr. E. Sullivan, in most of 

 which an imaginary Gilbert White is a prominent figure. 

 Mr. Herbert Railton's head- and tail-pieces are, for the 

 most part, delicate and attractive. As regards the notes, 

 Dr. Sharpe's name is, of course, a sufficient guaiantee 

 of the soundness of those on birds, and the only fault to 

 be found with them is that they are occasionally a little 

 wanting in succinctness and self-repression. Several of 

 Dr. Sharpe's colleagues at the British Museum have pro- 

 vided him with useful notes relating to their departments 

 of natural history, and a judicious selection has been 

 made from the notes of previous editors, especially Bell 

 and Harting. 



From what has been said above, it will be seen that this 

 is not really an edition of the book that White so care- 

 fully wrought into an artistic form, and that we all know 

 and love. It is not pleasant to say it, but said it must 

 be emphatically, that the liberties here taken with White's 

 work have absolutely no literary justification, and have 

 robbed it of much of that peculiar charm which, as Prof. 

 Newton has well said in his admirable article on White 

 in the " Dictionary of Biography," it is impossible to 

 explain in words. What would have been White's own 

 feelings if he had been forced to see in print the very 

 portions of his letters which, with his own good sense 

 and the respect of his age for publication, he had de- 

 liberately cut out, and the insertion of two hundred pages 

 of his gardening notes between the letters to Pennant and 

 those to Barrington ? If it be argued that (as Dr. Sharpe 

 seems to think) we learn something new about White 

 himself by getting an idea of the original form of his 

 letters and of the way in which he wrought his book out 

 of them, the plain answer is that we already know all 

 that is essential about him, and that one thing we know 

 for certain is that he had a sense of literary form which 

 has made his book immortal, and which should have 

 secured for it more reverential handling than is to be found 

 in these volumes. It might, indeed, be possibly justifiable to 

 print the whole of the original letters as they left his hand ; 

 but not as an edition of the " Natural History of Selborne," 

 which should always be allowed to stand exactly as his 

 genius designed it. It will be the duty of future editors 

 to see that none of the passages now interpolated are 

 NO. 1643, VOL, 63] 



allowed to creep permanently into the text of the original' 

 work. 



Dr. Sharpe's enthusiasm for his author is unquestion- 

 able, as may be seen from his brief but pleasant intro- 

 duction to the first volume ; so, too, is the labour that 

 he has spent on his editorial task. Bu t the perils of the 

 editor of a classic are great, and enthusiasm alone wilL 

 not teach him how to avoid them. 



OCIR BOOK SHELF. 

 The Romance of the Heavens. By A. W. Bickerton. 

 Pp. 284. (London : Swan Sonnenschein and Co., Ltd., 

 1901.) 55-. 

 The theory of constructive impact, of which a popular 

 account is given in the present book, appears to have had 

 its origin in an attempt to explain the phenomena of 

 new stars by the grazing collision of two dark bodies. 

 Hitherto the theory has not been hospitably received by 

 astronomers, and the more elaborate exposition now 

 presented will probably meet no better fate. The truth 

 seems to be that in spite of his claim to have discovered 

 numerous facts not known to " ordinary " astronomers, 

 the author lacks familiarity with spectroscopic work and 

 astronomical methods generally. He quite condemns 

 himself by suggesting (p. 235) that more confirmatory 

 evidence in the case of Nova Aurigae was only wanting 

 because astronomers, unguided by the theory, did not 

 make "more liberal and careful observation." As a 

 matter of fact, the most valuable records were photo- 

 graphic, and are still as much in evidence as during the 

 visibility of the Nova, and the observations certainly 

 cannot be interpreted as indicating the presence of three 

 bodies of the kind required by the theory. The theory 

 thus breaks down at the outset, and it would not be diffi- 

 cult to show the weakness of most of the " overwhelming "^ 

 astronomical evidence on which depends its extension 

 into collisions of nebulae, clusters and cosmic systems- 

 by which it is argued that the existing forms and distri- 

 bution of celestial bodies are completely explained. The 

 merest possibilities are frequently magnified into cer- 

 tainties, as, for example, the occurrence of variable stars 

 in pairs, and the preponderance of variability in double 

 stars. 



The resources of the theory appear to be unlimited. 

 While one collision produces a new star, another results 

 in a star cluster, another blows a planet into asteroids,, 

 and still another disperses a satellite into a ring such as 

 that of Saturn. 



We have not examined all the calculations which are 

 given, but we may point out that the results arrived at 

 for the separation of two stars of assumed distance and 

 velocity (p. 58) are in each case six times too great. 



The book is admirably written and is by no means 

 without interest ; but readers should be warned against 

 mistaking the author's assertions for demonstrated 

 truths. 



Les Diastases et leurs Applications. By E. Pozzi-Escot 



Pp. 218. (Paris : Masson and Co., 1900.) 

 This little volume forms one of the series of Aide- 

 Memoire, some of which have been previously reviewed 

 in this journal. Its modest preface disarms criticism, 

 " Le lecteur ne devra pas chercher ici I'expose de 

 theories nouvelles ; nous nous sommes contente d'exposer 

 les faits connus, de les relier I'un k I'autre et d'en tirer 

 chaque fois qu'il y a eu lieu des conclusions legitimes." 

 And further, " Forc^ment incomplet, nous esperons que 

 notre travail (qui n'est point fait pour des biologistes, 

 mais bien pour une collection destin^e aux ingdnieurs et 

 aux chimistes), rendra n^anmoins quelques services et 



