.\ %.v Notices respecting New Books, >' 315 



especially if, with regard to the numbers under 16^ 24 and 32, 

 we make some allowance for the approaching saturation of the 

 iron. 



I inferred from the experiments detailed in my last letter, that 

 little difference of attractive power would result from the mere 

 alteration of the shape of the sectional area of the iron of an 

 electro-magnet ; that view is confirmed by the experiments just 

 related, in which it will be seen that little difference exists be- 

 tween the magnetic powers of the first and second pairs ; and 

 even that difference may be partly accounted for by taking into 

 account the difficulty of winding the covered wire closely to the 

 surface of broad rectangular iron bars. 



The above magnets were wound to two thicknesses by the 

 covered wire, and in other respects were similar to those I before 

 used. The effect arising from increase of length may therefore be 

 estimated. These magnets, which were 30 inches long, wound 

 with eighty-eight yards of wire, and excited by a current of 6, 

 sustained a weight of 7000 grains at the mean distance of ^th 

 of an inch ; whilst the attractive power of the pairs marked VIII., 

 IX. and X., in my last, with the same electro-magnetic force, or 

 twenty-two yards of wire, and a current of 24, was 10646 grains. 



*f\ 5fC *^ ^ ^ 



I remain, dear Sir, 



Yours most respectfully, 



J. P. Joule. 



XL VIII. Notices respecting New Books. 

 L*Algehre(VOmarAlkhayydmi. Par F. Woepcke. Paris, 1851. 8vo. 



THERE is an old tradition that among the Arabic manuscripts 

 bequeathed by Walter Warner to the University of Leyden, 

 was one which treated of the algebraical solution of cubic equations* 

 In 1834 M. Sedillot discovered a manuscript fragment in the Royal 

 Library at Paris, which, it seemed probable, was part of the same 

 work ; of this he published some account {N. Jo. Asiat., May 1834 ; 

 Not. et ext. des MSS. de la Bibl. R., vol. xiii. pp. 130-136). M. 

 Libri afterwards found a complete manuscript in the Royal Library 

 {Histoire, &c., vol. i. p. 300), M. Woepcke has now published this 

 work, with the assistance of the fragment and the complete manu- 

 script just noted, and also the manuscript of the Leyden library itself. 

 As noted by M. Libri, the work does not contain the solution of 

 cubic equations, but only their geometrical construction, by aid of 

 the conic sections. But though the tradition which we have men- 

 tioned imposes upon us this depreciatory kind of description, it is 

 not the less to be noted that what we really have obliges us to form 

 a much higher idea of the Arabian algebra than could have been 



