October 13, 1921] 



NATURE 



205 



Baglioni. It gives a rapid summary of ethnology 

 and anthropology — good in its way, but, owing 

 to shortage of space, too compressed to be of 

 real value. The point is emphasised that modern 

 sociology shows a marked tendency to state and 

 solve its various problems in terms of physiology. 

 To Prof. Pembrey is due hearty congratulations 

 for the judicious and careful way in which he 

 has edited the volume. It must have required 

 endless patience and time. The translation is 

 good, and the number of actual errors detected 

 but few. The publishers have produced a book 

 of pre-war standard in paper, printing, illustra- 

 tions, and binding. It is a pity, however, that a 

 complete index for the five volumes was not in- 

 corporated in this concluding volume. 



E. P. C. 



Principles of Electrical Engineering-. 



Electrical Engmeering. By Dr. T. F. Wall. 

 Pp. xi -1-491. (London : Methuen and Co., Ltd., 



I921.) 215. 



ASL'RVEY of the principles of electrical engin- 

 eering intended for students in universi- 

 ties and the advanced classes in technical schools 

 is given in this book. The author's treatment of 

 the subject can be commended, although in places 

 the condensation will make it difficult for the un- 

 initiated to follow his reasoning. He begins by 

 a careful discussion of electrostatic theon,-, prov- 

 ing, in some cases by novel methods, the capacity 

 formulae which are used by engineers. He 

 describes how the dielectric is sometimes graded 

 in high-tension cables, and shows how the requisite 

 calculations to find the electric force in the 

 dielectric can be made. Xo mention is made, how- 

 ever, of the severe limitations imposed on the use 

 of intersheath methods of grading by the large 

 capacity current which flows in the sheath. The 

 corona effect is mentioned, but the formula given is 

 not so accurate as that due to F. W, Peek. The 

 formulae for the sparking voltages between 

 spheres are not given. 



On p. 194 it is stated that the standard values 

 for the resistance of copper at present in use are 

 those found by Matthiessen. This is not the case. 

 Electricians use the international standard of re- 

 sistance for copper given in Publication Xo. 28 

 of the International Electrotechnical Commission. 

 They also find it advisable to use three tempera- 

 ture coefficients : the "constant-mass" tempera- 

 ture coefficient, the volume resistivitv and the 

 mass resistivity temperature coefficients. 



On p. 202 Newton's law of cooling is given as 

 if it applied to radiation instead of to convection. 

 NO. 271 1, VOL. 108] 



It is deduced that the melting current of a fuse 

 wire varies as the i-5th p>ower of the diameter in- 

 stead of the I -2 5th power, which follows from 

 more accurate theory^. On p. 381, 1. 11, we take 

 it that "two-thirds" is a misprint for "three- 

 halves," as the capacity between two wires is 

 obviously increased by bringing a third wire into 

 the neighbourhood. 



The author attacks the problem of practical 

 harmonic analysis in the proper way. He takes 

 the Fourier solutions for a„ and b^, the coefficients 

 of the cosine and sine components of the Fourier 

 series, and computes their values by mathematical 

 quadrature. Taking m ordinates for the half- 

 wave, he writes : — 



2 r J"" 1 27r , , mn~\ 



a»= — y, cos«— -t-vjcosw — + . . . +y^cosn — , 



wL m m m J 



and a similar formula for ?>„. Taking fn.= io, 

 he finds the first, third, and fifth harmonics for a 

 given curve, and suggests that a similar analysis 

 will give the higher harmonics. It should have 

 been stated that more ordinates would have to be 

 measured if the higher harmonics are to be deter- 

 mined accurately. 



In our opinion the first and third harmonics are 

 best determined by dividing the base of the posi- 

 tive half of the wave into twelve equal parts and 

 then appMng Weddle's rule. To make reason- 

 ably certain of finding the fifth harmonic accur- 

 ately it would be necessary to divide the base 

 into eighteen, or better twent}--four, equal parts, 

 and draw the ordinates at the points of division. 

 If we apply the author's method to a rectangular 

 wave of height unity, we get 61=1-263 and 

 63 = 0-393. The true values, 1-273 ^"^ 0424, are 

 given by Weddle's rule. k. R. 



Semi-popular British Botany. 



.4 A'eit' British Flora: British Wild Floivers in 

 their Natural Haunts. Described by A. R. 

 Horwood. Vol. 3, pp. xi -f 25 1 -f plates 

 18-31; Vol. 4, pp. xi-H 257 -H plates 32-49; 

 Vol. 5, pp. xi-f 234 + plates 50-64; Vol. 6, 

 pp. xix-f232. (London: The Gresham 

 Publishing Co., Ltd., 1919.) 125. 6d. net each 

 vol. 



THE first two volumes of this work were 

 reviewed in Nature of April 21 last, p. 2^2. 

 Vol. 3 deals with flowers of the woods 

 and copses, roadsides and hedges, while the 

 fourth volume presents the flowers of "moun- 

 tains, hills, and dry places," "lakes, rivers, 

 ditches, and wet places," " waste places, gardens, 

 refuse-heaps, village greens, farmyards, etc." 

 While thoroughly unbotanical in that it scatters 



I 



