3<6 



NATURE 



[August '6, 1903 



Africa in 1898 was about 23 millions sterling. This 

 amount seems small for the huge continent, when we 

 reflect that in 190 1 the coal output of Wales alone was 

 worth 19^ millions, and that of Northumberland and 

 Durham about the same amount. But after reading 

 Prof, de Launay's book, it needs no prophet to predict 

 that Africa's mineral deposits will soon be more 

 largely utilised. 



ROWLAND'S WORK. 



The Physical Papers of Henry Augustus Rowland. 

 Collected for publication by a Committee of the 

 University. Pp. xi + 704. (Baltimore : Johns 

 Hopkins University Press ; London : Wesley and 

 Son, 1902.) Price 305. 6d. net. 



PROF. ROWLAND'S friends have been well- 

 advised in issuing as; a memorial to their late 

 colleague this volume of ihis collected papers. It 

 enables us to realise more fully all we owe to him 

 and to grasp the value and importance of his work. 



Commencing with an early note sent to the Scien- 

 tific American when the author was seventeen, the 

 list of scientific papers concludes with an article on 

 diffraction gratings, published in the new edition of 

 the " Encyclopaedia Britannica " after Rowland's 

 death. Then there follow some six addresses on 

 scientific subjects, a bibliography, and an account of 

 the dividing engines he designed. 



Dr. Mendenhall's commemorative address, delivered 

 shortly after Rowland's death, fitly forms an introduc- 

 tion to the whole, and gives us a glimpse of his life 

 and methods of work. 



Rowland's fame came to him early, though not 

 without some severe struggles and disappointments 

 on his part, and it is a satisfaction to us Englishmen 

 to know that it was Maxwell who first recognised his 

 genius. Prof. Mendenhall tells again the story of 

 his first serious paper, " On Magnetic Permeability 

 and the Maximum of Magnetism of Iron, Steel, and 

 Nickel," Phil. Mag., 1873. The paper was more than 

 once rejected in America because it was not under- 

 stood, and finally it was sent to Maxwell, who wrote 

 immediately that since the temporary suspension of 

 their meetings made It impossible to communicate the 

 paper to the Royal Society, he would send it to the 

 Philosophical Magazine, where it appeared in August, 

 1873, Maxwell having himself, to save time, corrected 

 the proofs. In this paper Rowland introduced the 

 idea of the magnetic circuit as the analogue of Ohm's 

 law, and developed the now well-known ring 

 method of measuring permeability. In 1875, on his 

 appointment as first professor of physics at the Johns 

 Hopkins University, he came to Europe and worked 

 for a time In Helmholtz's laboratory at Berlin, and by 

 his researches answered Tait's question, put to Max- 

 well in these words — ■ 



Will mounted ebonite disc 



On smooth unyielding bearing, 

 When turned about with motion brisk 



Nor excitation sparing, 

 Affect the primitive repose 



Of + and — in a wire? 



NO. 1762, VOL. 68] 



To which Maxwell replies — 



The mounted disc of ebonite ' 



Has whirled before nor whirled in vain, 



Rowland of Troy that doughty knight 

 Convection currents did obtain 



In such a disc of power to wheedle 



From its loved North, the subtle needle. 



And Maxwell goes on to explain that such convec- 

 tion currents will not produce electromotive force In a 

 neighbouring wire unless the speed of the disc were 

 variable, j 



The paper on the " Magnetic Effect of Electric 

 Convection," No. 12, in the volume before us, was 

 presented in the American Journal of Science for 

 1878; von Helmholtz had already announced the re- 

 sult to the Berlin Academy in 1876. Rowland re- 

 turned to the problem with the same result In 1889, In 

 a paper presented In the Philosophical Magazine, No. 

 43 of his collected works. As Is well known, the 

 results were challenged by Cr^mleu shortly before 

 Rowland's death. Many readers of Nature will 

 remember the interesting occasion in Section A of 

 the British Association at Glasgow, when Cr^mieu 

 described how he had failed to obtain the effect. 

 Those present felt that In view of the confirmation 

 of Rowland's results obtained at Baltimore by Pender, 

 Cr^mleu must have been misled, but no one could 

 put his fingers definitely on the error. It. is satis- 

 factory to know, from the recently published joint 

 work of Cr^mieu and Pender, that Rowland was 

 right, and that a convection current of electricity does 

 produce a magnetic field. 



The research Into the value of the British Associa- 

 tion unit of resistance. No. 15, and a determination 

 of the value of " v,'' No. 44, complete the series of 

 fundamental electrical researches, though his collected 

 papers contain many other memoirs of real import- 

 ance. 



In his experiments on the absolute unit of resist- 

 ance, Rowland shows his usual acumen as a critic 

 and skill as a mechanic and observer. Various lines 

 of argument had shown that the B.A. unit, supposed 

 to represent 10^ C.G.S. units of resistance, was in 

 error. Rowland sums up effectively his criticisms on 

 the method of the B.A. committee and points out the 

 sources of error in Weber's method by damping 

 adopted by Kohlrausch. He then describes his own 

 method, a modification of that originally proposed by 

 KIrchhoff, and, after a careful account of his ap- 

 paratus and measurements, arrives at the result 

 I B.A. unit = 0.991 1 X 10' C.G.S. units. A repetition 

 of his experiments in 1884 gave 0.98627, while about 

 the same time his pupil, Kimball, using Lorenz's 

 method, arrived at the result 0.9864. The value 

 obtained at the Cavendish Laboratory was 0.9867. 



Part ill. of the collected papers deals with the work 

 on Heat, and foremost among these Is the great 

 memoir on the "Mechanical Equivalent of Heat," a 

 work which, if it stood alone, would have made 

 Rowland's name as the foremost physicist of his 

 nation. 



The refinements of modern thermometry have 

 enabled us to Introduce some small corrections into 

 certain of the results, but the work remains unrivalled. 



