870 Prof. H. L. Callendar on 



length. That is, they contract in a particular azimuth only, 

 retaining in the azimuth normal to this the diameter of the 

 un stretched wire. The cross section of a wire behaving in 

 this way appears to be roughly elliptical, so the contraction 

 is approximately uniform across the wire. The minor axis 

 oE such a cross section is, in the case of tin, often only 0*4 of 

 the major. The ring-shaped markings appear on the flatter 

 surfaces of the wires, which are shown in the photographs ; 

 they are similar on each flattened side and run together at a 

 sharp angle at the edges of the flattened wire. The plane in 

 which the flattening takes place must be determined by 

 chance asymmetrical irregularities, for with tin two or more 

 •such different planes sometimes occur on the same wire, 

 separated by small lengths in which the wire retains its 

 original circular form, and shows no regular markings. 



Thus on extending wires of the soft pure metals mercury, 

 iin, and lead (and also sodium and potassium, as shown by 

 Baker), we are able to get surface markings of great regu- 

 larity, accompanied by a contraction of the wire in one 

 particular direction only. The markings present the appear- 

 ance of a series of equal layers which have been sheared over 

 one another, as would be the case with a half cylinder com- 

 posed of semicircular plates if the plates were all tilted over 

 to make an acute angle with the axis of the cylinder. They 

 ;are probably due to large uniform crystals, of a size com- 

 parable with the diameter of the wire, arranged in la} r ers, 

 which behave somewhat in the way suggested. It is notice- 

 able that the metals which give the phenomenon are all very 

 soft, a condition to which large crystals are known to be 

 favourable. The purer lead which shows the markings is 

 softer than ordinary commercial lead. The phenomenon has 

 evidently nothing to do with the processes to which the wire 

 is subjected during manufacture, as the specimens of lead 

 and tin have to be thoroughly annealed if they are to show 

 it well. 



XCVI. Thermodynamics of Radiation. By Hi. L. Callendar, 

 M.A., LL.D., P.R.S., Professor of Physics at the Imperial 

 ■College, S. W* 



IN the number of this Journal for October 1913, p. 787, I 

 gave a brief sketch of a theory of radiation and specific 

 heat, which appeared to be worth recording on account of its 

 simplicity and its good agreement with experiment. The 

 formula given for the distribution of energy in full radiation 

 * Communicated by the Author. 



