i'-Z liORD RAYLEIGH. ON THE STRESSES IN SOLID BODIES &f,'. 



is examincd tlirougli the thickness of tlie slieet, a great waiit of uiiifor- 

 mity is manifestée!. lu spite of the shortness of the distaiice traversée!^ 

 there is in phaces considérable revival of light with intermediate irregu- 

 larly disposed dark bands. The course of thèse bands is altered wheu 

 by fracture any part is relieved froni the constraining influence of neigh- 

 bouriug parts. To make an exaniinatiou by liglit transiuitted edgewise 

 it was necessary to immerse the glass in a li([uid of nearly equal refrac- 

 tivity (beuzole with a little bisulphide of carbon) contained in a small 

 tank. The width^ traversed by tlic light^ was about 1 cm. In this wàj, 

 and with the aid of a magnifterj the condition of the varions hiyers 

 could be well made out. The dark bands of no double refraction seemed 

 to be nearer to the faces than according to the calculation niade above^ 

 but the whole thickness is so small that this observation is scarcely to 

 be relied upon. The interior was in a state of tension^ and the double 

 refraction was nearly sufficient at the middle to give the yellow or brown 

 of the tirst order. By the action of hydrofluoric acid ou the lower end 

 of one of the strips the outermost layers were dissolved away. This cau- 

 sed a drawing together of the dark bands towards the middle, and though 

 a good deal remaiued the light was nuicli reduced. 



The cause of the louglii'ii'Dicj lias been sought in a spécial crystalliue 

 condition due to the sudden cooling. There may be something of this 

 nature; but it would seem that most of the peculiarities manifested 

 may be explained by référence to the known condition of stress. T])e 

 fracture of glass is usually due to bending, and the failure occurs at 

 the surface which is under tension. If, initially, the superticial layers 

 are under strong compression, a degree of bending may be harmless 

 which otherwise would cause fatal results. It seems possible also that 

 the superficial compression may be the explanation of the spécial hard- 

 ness observed. 



A short length of glass rod in its natural imperfectly annealed con- 

 dition may be used to illustrate syuimetrical stress. The ends may be 

 grouud, and either polished or provided with cover-glasses ceinented 

 with Canada balsam. In the spécimen examined by me the colours varied 

 from the black of the tirst order ou the axis to the red of the second 

 order near the surface. The length of the cylinder was 1.6 cm. and the 

 diaiueter 1.8 cm. 



