171 



alloys of various metals, using equivalent proportions, and determined 

 their conducting powers. The general result obtained is, that alloys 

 may be classed under the three following heads : 



1st. Alloys which conduct heat in ratio with the relative equiva- 

 lents of the metals composing them. 



2nd. Alloys in which there is an excess of equivalents of the worse 

 conducting metal over the number of equivalents of the better con- 

 ductor, such as alloys composed of iCu and 2Sn ; iCu and 3Sn; 

 iCu and 4Sn, &c., and which present the curious and unexpected 

 result that they conduct heat as if they did not contain a particle of 

 the better conductor ; the conducting power of such alloys being the 

 same as if the square bar which was used in the experiments were 

 entirely composed of the worse conducting metal. 



3rd. Alloys composed of the same metals as the last class, but in 

 which the number of equivalents of the better conducting metal is 

 greater than the number of equivalents of the worse conductor ; for 

 example, alloys composed of iSn 2Cu; ISn 3Cu ; iSn 4Cu, &c. ; 

 in this case each alloy has its own arbitrary conducting power, and 

 the conductibility of such an alloy gradually increases and tends 

 towards the conducting power of the better conductor of the two 

 metals composing the alloy. 



Experiments were also made with bars composed of various metals 

 soldered together, in order to compare the results obtained with alloys 

 with those afforded by the same metals when mixed. 



The first part of the paper concludes with the conducting power 

 of several commercial brass alloys. 



The second part, which will shortly be published, will contain the 

 conduction of heat by amalgams. 



II. " On the Surface which is the Envelope of Planes through 

 the Points of an Ellipsoid at right angles to the Radius 

 Vectors from the Centre." By ARTHUR CAYLEY, Esq., 

 F.R.S. Received February 22, 1858. 



(Abstract.) 



The consideration of the surface in question was suggested to me 

 some years ago by Professor Stokes ; but it is proper to remark, that 

 the curve which is the envelope of lines through the points of an 



