58o 



NA TURE 



[October 12, 1905 



have done their work. In issuing a register of zoo- 

 logical work for 1904 so early as September of the 

 present year, the editor and publisher have indeed 

 beaten our own "Zoological Record"; but it must 

 be remembered that in the present volume is included 

 a considerable amount of literature belonging to 

 earlier years, while it is difficult to believe that the 

 whole of the papers for 1904 can be included. 



It might be imagined, for those not conversant 

 with the two works, that the " Zoologischer Jahres- 

 bericht " is a serious rival to the " Zoological 

 Record," and that the publication of the one renders 

 that of the other superfluous. As a matter of fact, 

 this is not the case; for, in the first place, it is highly 

 desirable that a record of zoological literature should 

 be published in English, and, in the second, the 

 two publications do not cover the same ground. The 

 " Zoological Record," for instance, is specially de- 

 voted to the systematic aspect of the subject, particular 

 pains being taken to include the names of all new 

 species and subspecies. In the Continental work, on 

 the other hand, systematic work is rigorously ex- 

 cluded, and attention concentrated on the bionomical, 

 anatomical, and phvsiological aspects of the subject. 

 The two records are therefore to a considerable extent 

 supplemental and complementary to one another, 

 more especially as in the one before us a somewhat 

 full precis of the main subjects of the more important 

 papers forms an important feature. The practice of 

 including all the papers on Vertebrata under a single 

 heading does not, indeed, appeal to us; but then, it 

 is true, this is in some degree compensated by 

 dividing the summary of their contents into their 

 respective class-positions. So far as we have been 

 able to judge, the quotations of the titles of the papers 

 and the references to their places of publication are 

 singularly free from error, and the volume, like its 

 predecessors, cannot fail to be of the highest value to 

 all workers in morphological and anatomical zoologv. 



R. L.' 



Examples in Arithmetic. By C. O. Tuckey. Pp. 



xii-l-241 + xxxix. (London: George Bell and Sons, 



1905.) Price 3s. 

 The Primary Arithmetic. Parts i. and ii. Edited by 



Dr. Wm. Briggs. Pp. 80 and 94. (London : The 



L'niversitv Tutorial Press.) Price 6d. each. 

 These books are intended for the use of teachers 

 who instruct their classes orally in the processes 

 and rules of arithmetic, and who only require the 

 assistance of graduated sets of exercises. In the 

 work by Mr. Tuckey the course is fairly complete, 

 embracing the usual commercial arithmetic, with 

 a chapter on the application of proportion to prob- 

 lems in geometry and physics, and a section de- 

 voted to numerical computations by the aid of com- 

 pound interest, logarithmic and trigonometrical tables, 

 in which a little elementary trigonometry is intro- 

 duced. There are examples on graphs and squared 

 paper work, and the users of the book will have an 

 abundant choice of exercises of modern type. 



" The Primary .\rith_metic " will be complete in 

 three parts. The first part gives sets of e.xercises 

 on the four simple rules and on the compound rules 

 for money. The examples increase in difficulty by 

 almost imperceptible stages, beginning with those of 

 the sirhplest kind, and they are suitable for verv 

 yoimg scholars. Part ii. completes the compound 

 rules for weights and measures, including the metric 

 system. Then follow exercises on vulgar fractions 

 and on practice and invoices. In these two parts, as 

 well as in the book by Mr. Tuckey, the answers to 

 the exercises occupy a considerable space at the end 

 of each volume. 



NO. 1876, VOL. yi] 



LETTERS TO THE EDITOR. 

 [The 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 otlier part of "Satuke. 

 A'o notice is taken of anonymous communications. \ 



A Magnetic Survey of Japan. 



In Nature of .April 20 (vol. Ixxi. p. 578), Prof. .A. 

 Schuster has given a comprehensive review of the magnetic 

 survey of Japan with a friendly criticism. The responsi- 

 bility of its writer may be a sufficient excuse for the 

 following remarks partly in way of reply. 



Prof. .Schuster directs attention to the small space given 

 to the description of the working of the instruments. This 

 arises from the fact that these instruments were essentially 

 the same as the one used in the previous survey of 1887, 

 and described in vol. ii., pp. 178-193, of the Journal of 

 the College of Science, Imperial University, Tokyo, to 

 which the reader is referred for details. A few improve- 

 ments that have since been made are mentioned in the 

 present report, pp. 7-8. 



We are glad to see that the methods adopted for calcu- 

 lating the corrections for heights of stations and the way 

 of disposing with the vertical current met his approval ; 

 only Prof. Schuster seems to attribute these currents to 

 uncertainties in the observations, whereas we infer that 

 they are as much, if not more, due to the inadequacy of 

 the empirical formulae, from the fact that they vanish near 

 the middle of the several countries treated (p. 125). 



Perhaps the more important point is with regard to the 

 question of the seat of action. To avoid confusion, it 

 might be well to remark that the word potential is used 

 in different senses by different writers; some use it to 

 denote a function which satisfies the Laplacian equation 

 V"V = o, and others to denote the line integral of any 

 irrotationally distributed vector, whether the solenoidal 

 condition be satisfied or not. It is in the latter general 

 sense that the word is used in the report. 



Now Gauss's method of separating internal and external 

 sources of action is based upon the assumption that these 

 sources are entirely separated from each other by a free 

 space ; in other words, the Laplacian equation holds strictly 

 over a finite portion of the space surrounding the earth 

 surface. This is very plausible when we consider the earth 

 as a magnetised body, as appears a posteriori. But when 

 we abandon the restriction of the solenoidal distribution 

 the method is no more applicable, and the observation of 

 force over a spherical surface is not sufficient to settle the 

 seat of action, although it may be expansible in harmonic 

 form if its distribution is continuous, so that the Gaussian 

 expansion must be taken in " Gauss's sense " (end of first 

 paragraph, p. 140 of the report). 



The possibility of the distribution of magnetism in the 

 space surrounding the earth surface might appear quite 

 extravagant, and may be included amongst what Gauss 

 calls " boldenlose Phantasien," so long as we are con- 

 sidering the main causes of the terrestrial magnetism ; but 

 when we come to discuss the external causes and the 

 horizontal atmospheric current the effects of which amount 

 to only a small fraction of the observed forces, our 

 assumption of the distribution being thoroughly solenoidal 

 would seem subject to doubt, or at least to require observ- 

 ational evidence, so that " strictly speaking, the mode 

 of distribution must remain perfectly arbitrary so long as 

 we adhere solely to the observed elements "of magnetic 

 forces on a spherical surface, when no further assumption 

 than the Newtonian law of action is admitted. 



It may not be unnecessary to add here that the search 

 for the seat of action from observations of force over a 

 surface is an inverse problem, and includes any arbitrarv 

 distribution of magnetism the resultant effect of which 

 vanishes on that particular surface ; we can put any system 

 of magnets or electric circuits outside the surface,' pro- 

 vided we envelop that surface with a counteracting shell 

 or shells over which a proper distribution of magnetism 

 is made according to Green's method of finding the 

 density of induced electricity on a conductor, besides anv 



