Astronomical and Nautical Collections. 393 



of the different lines of- agreement or discordance, we must 

 fix our attention on the point we wish to observe, by bring- 

 ing to it the wire placed in the focus of the lens of the 

 micrometer; or, what is still better, by substituting for this 

 wire a screen furnished with a narrow slit, which only allows 

 the light of this part of the fringe to pass through it. The 

 horizontal or vertical polarisation of the points of complete 

 agreement or disagreement ceases to take place when we 

 intercept one of the pencils, and receive in the slit the light 

 of the other only: it is then polarised like this latter alone; 

 that is to say, in a direction inclined in an angle of 45° to 

 the horizon. Hence it follows, that the polarisation in the 

 primitive plane, or at the azimuth 2 f, results from the union 

 of the two pencils, and does not take place in either of the 

 pencils taken separately, which are always polarised in the 

 direction parallel or perpendicular to the principal section, 

 whether we observe them separately by the lens, or together 

 without the lens, so as to see both luminous points at once, 

 and observe both their polarisations. The lens preventing 

 the distinct vision of the two points by the expansion of their 

 images, which mixes their rays at the bottom of the eye, 

 reproduces there the interferences which took j^lace in the 

 focus; and for this reason it is necessary for viewing the 

 phenomena of interference, when the two images of the lumi- 

 nous point are not mixed together, or, in other words, when 

 the two systems of undulations, which interfere, form a sen- 

 sible angle with each other. It is easy to be convinced that 

 the lens has no other effect here, and that it does not exer- 

 cise any sensible polarising influence, by looking at a lumi- 

 nous pencil polarised in a known direction, for it will be 

 seen that the interposition of the lens makes no change in it. 

 The polarisation, therefore, that we have observed, in the 

 primitive plane, and at the azimuth 2 i, depends wholly on 

 the combination of the two pencils emerging from the rhom- 

 boids which cross each other. 



If, leaving the principal sections always perpendicular to 

 each other, we turn the two rhomboids round, we may 

 remark, in every position of the system, that the lines of the 

 fringes, which answer to a difference in the paths of an even 

 number of semiundulations, are polarised in directions pa- 



