temporarily produced in Isotropic Bodies. 351 



presents itself infallibly whenever we make use of thin plates ; it 

 is impossible to render the surfaces between which the pressure 

 is exerted so completely plane and parallel that they shall rigor- 

 ously produce nothing more than an effect of compression. It 

 is therefore necessary that the glass itself offer sufficient resist- 

 ance to flexure to cause its effect to become insensible, and this 

 is attained by giving the plate a thickness (height) of from 2 to 

 3 centimetres. It is hardly possible to obtain by refoulement so 

 thick a glass ; on the other hand, the cast glasses always present 

 a great number of striae in a direction perpendicular to their 

 pressure, which is the precise direction in which they would be 

 traversed by the light. I have had recourse to the juxtaposition 

 of several layers of plate glass placed one upon the other; 

 these layers, after having been united by Canada balsam, have 

 been cut in such a manner as to form only a single thick plate, 

 perfectly transparent, and which fulfills all the required condi- 

 tions. It is easy to ascertain the dimensions which it is neces- 

 sary to give to such a plate, to render it fit to measure a force of 

 any given magnitude. Plate glass supports a pressure of 3 kilo- 

 grammes per square millimetre ; a square plate of 20 centi- 

 metres the side, could therefore, without inconvenience, be sub- 

 mitted to a pressure of 120,000 kilogrammes. But long before 

 it attains this limit the two images would become completely 

 white, in consequence of the great difference of path between the 

 two rays. 



In fact, according to our table, the greatest difference of path 

 which could be made use of is 0*002 of a millimetre ; the charge 

 corresponding is given by the formula 



^ 0-002x6180x200 .on^oii 



r= fvToT = 12,942 kilogrammes, 



or in round numbers, equal to 13,000 kilogrammes, a limit 

 which could not be exceeded without employing a larger plate. 



But without having recourse to a plate of larger dimensions, 

 which would at the same time diminish the sensibility of the 

 instrument, we can augment its power by a very simple artifice. 

 Suppose that we have arrived at a pressure of 13,000 kilogrammes; 

 the extraordinary ray would be in advance of its ordinary a di- 

 stance of 0*002. Let us now place in the line of the rays between 

 the two prisms (at /, fig. 3), a plate of quartz cut parallel to the 

 axis, and of such a thickness that of itself it could produce a differ- 

 ence of path of 7X, or of 0"004. The quartz is a positive crystal ; 

 and hence by adding the plate to the compressed glass, we change 

 the sign of the difference of path d; — from —0*002 it becomes 

 + 0*002. The colours of the two images have therefore suffered 

 no change ; but by now augmenting the pressure, they will gra- 



