546 



NA TURE 



[OCTOHER 8, 1896 



the subject to mere exercises in algebra and geometry ; 

 the former is the only process for acUancing true know- 

 ledge. It is unfortunately the mathematician's trans 

 former that figures so largely in the present volume. 

 The subject of iron testing, the magnetic qualities of 

 iron, its selection, and the effects of use on it for trans- 

 former cores, though fundamental matters of principle in 

 the case of the real transformer, are not so much as 

 mentioned, in spite of all that has been lately done in 

 this matter. The avoidance of eddy current losses in 

 the copper circuits and frames, the effect of magnetic 

 leakage. in causing such copper eddy current losses, the 

 processes of ventilating large transformers, and the real 

 difficulties of insulation, and the specialities of design 

 for various purposes, are not named. The practical 

 man, looking for approved principles of design in the 

 case of the transformer, asks for the bread of practical 

 •experience ; he is here presented with the stones of 

 an artificial theory. 



OUR BOOK SHELF. 



Mechanics for Beginners. By Linn^us Gumming, M.A. 

 Pp. viii -I- 247. (London : Rivington, Percival, and 

 Co., 1896.) 



Twenty years ago a Committee of the British .'Associa- 

 tion recommended that the school teaching of physics 

 should begin with a course of elementary mechanics, 

 treated from an e.xperimental point of view, and the 

 ■opinion expressed in the Physics Section of the Associa- 

 tion this year was in support of that view. Mr. Gumming 

 has for some years been endeavouring to act upon the 

 recommendation in the science classes of Rugby -School, 

 and the jM-esent book contains the course which his 

 •experience has proved to be the most suitable for 

 beginners. 



The book does not begin with dynamics, for though 

 l\Ir. Gumming recognises the scientific advantages which 

 this subject offers to the study of mechanics, he has 

 found it too abstract for yonng students. Statics lends 

 itself to experimental treatment, and is able to appeal 

 ■directly to the convictions and interests of boys beginning 

 the study of science. The first part of the book is, 

 therefore, devoted to this branch of the subject, dynamics 

 being treated in the second part, and hydrostatics in 

 the third. 



Teachers and students who are familiar with the 

 author's books on electricity and heat, will know the 

 ■character of his work. The present volume is thoroughly 

 practical, is very clearly illustrated, and will doubtless 

 "find its way into many schools. It shows how mechanics 

 may be experimentally taught in schools, and the prin- 

 ciples demonstrated with simple apparatus ; it thus 

 contains the elements of a sound scientific education. 



We regret to note the absence of an index, for no 

 text-book is complete without one. 



Hints on Elementary Physiology. By Florence A. Haig- 

 Brown. Pp. xii -F 121 ; 20 illustrations. (London : 

 J. and \. Churchill, 1896.) 

 These "Hints" are based upon notes taken by the 

 authoress and her sister while attending lectures and 

 demonstrations gi\en to probationers at St. Thomas's 

 Hospital. They will be found helpful as a means of 

 giving a general idea of the functions of the \ arious parts 

 of our bodies ; and nurses who read them will acquire 

 knowledge which will lead to the intelligent performance 

 •of duties. 



NO. 1406, VOL. 54] 



LETTERS TO THE EDITOR. 



[ The Editor does not hold himself responsible for opinions ex- 

 pressed by his correspondents Neither can he undertake 

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

 inanitscripts intended for this or any other part of NATURE. 

 No notice is taken of anonymous commiinications.'\ 



The Utility of Specific Characters. 



I AM anxious to avoid prolonging this discussion, but I should 

 like to say a word in answer to I'rof. Pearson and Mr. Cunning- 

 ham, The letters of these gentlemen may lead some readers to 

 infer that I have only just recognised the hypothetical nature of 

 the law of growth of crabs assumed in the report of last year, to 

 which reference has been made. I wish, therefore, to point out 

 that the hypothetical nature of this assumption is recognised in 

 the report itself (Koy. Soc. Proc, vol. Ivii. pp. 367-368), and in 

 the remarks which accompany it (ibid., p|). 38 1-382). 



So far as I remember, this point was dwelt upon by at least 

 one speaker in the discussion which followed the reading of the 

 report : and the fact that I have spent the whole of my spare 

 time, since the report was read, in an endeavour to ascertain the 

 actual law of growth, is evidence that I have not been blind to 

 its importance. 



I must ask Prof. Pearson's leave to postpone a discussion of 

 the actual law of growth until I have worked out the results of 

 all my observations. 



As for Prof. Pearson's second point — that correlations may 

 arise by chance— I altogether fail to follow him ; and the data 

 which he gives concerning his hypothetical breed of cows, do 

 not seem to me sufficient to .serve as the lusis of further dis- 

 cussion. \V. F. R. Wki.hon. 



Fnisersity College, London, October 3. 



An Error Corrected. 



We regret to have to acknowledge a mistake which we have 

 made in a communication to the Paris Academie des Sciences, 

 reproduced in Nature of Aug. 27. It refers to the densities of 

 helium ; it does not affect the experimental results, nor the con- 

 clusion that helium has been split into two portions of unequal 

 density ; but it affects the figures assigned to these densities. 

 The hypothetical case was stated that a mixture of four volumes 

 of oxygen and one volume of hydrogen would diffuse in equal 

 times, and therefore could not be separated. This conclusion is of 

 course wholly wrong, and likewise in conseijuence the densities 

 calculated for helium on a similar supposition. The densities of 

 the two fractions of helium are therefore those found experi- 

 mentally, viz. I 874 and 2 133. It is right to observe that these 

 figures stand for densities caf.ulated from the observed rates of 

 diB'usion, and not from direct weighings 



Wii.i.iAM Ramsay. 

 J. XoRMAN Collie. 



University College, London, Octolier I. 



The Departure of the Swallows. 



I l)(i )ioi know whether the eccentric behav'our of the swallows 

 tliis year is of sufficient interest to justify me in troubling you 

 with a letter. I am not the only person in this part of the world 

 whose attention it has attracted. 



Everybody is familiar with the spectacle of large assemblages 

 of swallows immediately preceding their total disappearance ; 

 usually, I think, in early October. This year great multitudes 

 were assembled here on September i ; both Hying about this 

 house, and at perch on ruils and telegraph wires. On the two 

 next da)"s only one or two were \"isible, and on the two days 

 succeeding none at all. I concluded that they h.id antedated 

 their departure by a month, although in this locality the steady 

 sunshine and dryness had not then ceased ; but on Sept. 6 large 

 numbers appeared, to disappear again the next morning. Since 

 then their action, or the action of some swallows, has varied 

 nearly in accordance with the twelve day.s' account which is 

 appended. Yesterday (September 30) none were visible, nor are 

 any to-day. But there has been no large assemblage immediately 

 previous. 



II you pay any attention at all to these remarks, please to 



