Precision Method of uniting Optical Glass, 277 



In addition, cements frequently strain the materials which 

 they hold together to such an extent that their optical pro- 

 perties are considerably altered. Fig. 1, PL V., illustrates 

 the distortion of a glass plate 1*25 in. across, 0*1 in. thick, 

 cemented to a piece of a brass tube by a strong and very 

 well-known adhesive. The interference-rings are produced 

 between the plate and a proof-plane of quartz ; the plate was 

 originally very flat, and showed less than one interference- 

 ring across its entire face. After cementing, the series of 

 rings appeared, indicating a depression in the centre of the 

 plate of about -^qqq in. There appears to be considerable 

 force acting radially inwards ; such distortion would render 

 the glass plate useless for almost any accurate optical instru- 

 ment, especially if polarized light were passed through it or 

 interference-phenomena were being observed. 



Glass surfaces can be polished without difficulty so nearly 

 flat — or so nearly of the same curvature — that, when carefully 

 cleaned and freed from dust, they show less than one inter- 

 ference-fringe per inch when placed together and examined 

 by light reflected from the interfaces. Such surfaces can 

 be pressed into optical contact by the application of pressure 

 of several pounds per square inch, and the pieces of glass 

 act as one for optical purposes : for example, if the hypo- 

 tenuse of a 45° prism is in such contact with another piece 

 of glass, no total reflexion — indeed, no reflexion at all — is 

 obtained from the common surface. They can easily be 

 separated by sudden warming or cooling, by force, if applied 

 in a 'peeling manner, and by liquids ; the latter gradually 

 creep in and lift the surfaces apart. The joint is very 

 resistant to forces tending to slide one surface over the 

 other. 



A cell constructed of two plates of glass in optical contact 

 with a U-shaped separating piece would have the advantages 

 of great accuracy of parallelism between the windows ; very 

 little strain in the windows if the components were of one 

 kind of glass and put together at the same temperature ; 

 and an "all glass" construction, involving no cement to con- 

 taminate liquids placed therein, and to perish eventually 

 under the action of solvents or corrosive substances. But it 

 would be quite useless, owing to the ease with which such 

 surfaces can be separated by the methods described above. 



It is well known that if glass is heated it gradually becomes 

 less viscous as the temperature rises, passing through all 

 degrees of viscosity until at a high temperature it is quite 

 mobile ; there is thus no sharp melting-point, but only an 

 approximate position which varies from about 1000°-1400° C. 



