September 23, 1915] 



NATURE 



87 



writings of history under the conception of 

 "Deutschland iiber alles." And not only the 

 writing- of history, for it seems that German edi- 

 tions of Caesar's Commentaries have been sub- 

 jected to surg-ical treatment to make them con- 

 formable with the axiom. It is not to be supposed 

 that Prof. Duhem has not respect for g-iants like 

 Leibniz and Gauss; indeed, in the fourth lecture 

 and the appendix, he gives (more philosophically 

 than in the rest of the book) careful appreciation 

 of the geometrical spirit which is at once the 

 strength and the weakness of German science. 

 Of this geometrical spirit English science is all 

 but destitute ; its strength is in intuition. The 

 characteristic French qualities are order and clear- 

 ness, and I'esprit de finesse ; and each has much 

 to learn from the others. 



Mineral Resources of Minas Geraes (Brazil). By 



A. F. Calvert. Pp. xvi+ioo + 127 plates. 



(London : E. and F. N. Spon, Ltd. ; New York : 



Spon and Chamberlain, 1915.) Price 65. net. 



This little volume contains a eulogy of the 



mineral wealth of the famous State of Minas 



Geraes, the importance and value of which have 



long been well known to all who have studied, 



however superficially, the distribution of the 



> world's mineral resources. The author appears to 



I have searched diligently all available records of 



Brazilian mining and has evidently read widely 



and extensively ; it is only a pity that he did not 



supplement his want of knowledge of technical 



i matters by getting- some competent mining- en- 



I gineer to revise his proof sheets. This would 



^ have saved him from repeating such a foolish 



f criticism as that of Sir Richard Burton upon the 



shortcomings of the Brazilian miner : — " The 



Davy and the Geordie were equally unknown to 



him," the mines referred to being gold mines ( ! ) ; 



it would have prevented him from persistently 



writing " phosphorous " instead of " phosphorus " ; 



it would have most certainly corrected his version 



of the reports of Baron von Eschwege, from which 



it seems that the author is unaware that the 



German equivalent of " blast furnace " is " Hoch- 



ofen," and that "high furnace" is a phrase that 



conveys no meaning- to the British smelter. 



Perhaps the most interesting sentence in the 

 book is one in the preface, where the author sum- 

 marises the causes that have led to the poor 

 results obtained in Brazil, as a rule, from such 

 i^reat natural resources; he enumerates them as 

 follows: — "Liberation of the slave workers, bad 

 legislation and exorbitant taxation, lack of railway 

 communication, political unrest and financial in- 

 stability, and the incompetence and dishonesty that 

 have attended the exploitation of the mines and 

 : the management of the companies that have been 

 jj foiTmed to work them," and goes on to point out 

 that in his opinion these drawbacks are being 

 steadily removed. Most British capitalists who 

 have had experience of extensive business opera- 

 tions in Brazil would be disposed to add several 

 ^■ery significant items to the above list, not the 

 least important of which would be the rapid 

 fluctuations in the rate of exchange. 

 1^ NO. 2395, VOL. 96] 



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 other part of Nature. No notice is 

 taken of anonymous communications.] 



Magnetic Measurements. 



With reference to the very kindly review in Nature 

 of July 29 of " Electrical Instruments in Theory and 

 Practice," by Mr. U. A. Oschwald and myself, your 

 reviewer takes exception to the suggestion on p. 32 

 and elsewhere that "in view of its great simplicity 

 the method of testing by the magnetometer ought to 

 be more used than at present." 



The reason urged for the rejection of magnetometer 

 methods is that disturbances from neighbouring 

 current-carrying conductors would render it useless, 

 and presumably ballistic, or other methods should be 

 used instead. 



I wish to point out on behalf of Mr. Oschwald and 

 myself that we did not suggest the adoption of 

 magnetometer methods in any haphazard way. Both 

 of us have had considerable experience of magnetic 

 work, and the suggestion was made for the follow- 

 ing reasons 



(i) In ballistic work, if a Thomson moving needle 

 galvanometer is used, as is frequently the case, the 

 same sources of error are present. 



(2) No ballistic galvanometer can be calibrated very 

 accurately at present. The values of the constant as 

 determined by Messrs. Hadfield and Hopkinson 

 (Journal Inst. Electrical Engineers, vol. xliv., p. 270) 

 by condenser, standard field, and steady deflection, 

 differ amongst themselves by about i per cent. ; and 

 there are generally other sources of error in mag- 

 netic work. 



Novv, when electric light wires are laid in a build- 

 ing, the flow and return wires are generally close 

 together in casing, or iron tubing, so that the effect 

 with any reasonable precautions is less than i per 

 cent. Hence the method is practically equal to the 

 ballistic one, so far as accuracy is concerned 



Again, for rapidly comparing specimens of iron by 

 the zero method, the objection vanishes, and there is 

 nothing to approach the magnetometer for this pur- 

 pose. 



In my opinion, the ballistic galvanometer is a con- 

 siderably overrated instrument for this purpose, due 

 possibly to the fact that errors in its constant are 

 conveniently forgotten in the course of research. 



Apart altogether from accuracy, there is the out- 

 standing feature of the simple magnetometer, viz. 

 that it measures directly the forces we are dealing 

 with, whereas the ballistic method depends on measur- 

 ing another quantity altogether. I think I am not 

 wrong in hoping that the magnetometer will be 

 partially restored to its ancient position as the premier 

 magnetic testing instrument. 



W. H. F. Murdoch. 



September 3. 



With reference to Mr. Murdoch's letter on the 

 magnetometer method of testing iron, I described 

 it as an admirable method under suitable con- 

 ditions. I tried the method about three years ago 

 in my laboratory, which is close to a track return 

 tramway line, and had great trouble with it owing 

 to zero fluctuations. The same difficulty has been 

 found in other laboratories close by, where instruments 

 with very weak magnetic control have been set up. 

 I did not take exception to the suggestion that the 



