FOR SOLAR OBSERVATIONS. 29 



series of angles between 41° 41' and 79 c 11'. When the light 

 was transmitted through one plate, it ought to have been pola- 

 rised, according to the preceding law, at an angle of 88° 38'; 

 but upon examining the image with a prism of calcareous spar, 

 it did not vanish in any position of the spar, even when the 

 incidence was greater than 88° 38'. A similar aberration from 

 the law takes place when the pencil is transmitted through 

 owe, two, three, or four plates. 



In order to discover the cause of this anomaly, I made a 

 great variety of experiments, without obtaining any satisfac- 

 tory result ; and it was only by examining the phenomena of 

 the polarising eye-piece that my attention was again called 

 to the subject. We have already seen, that the pencils 

 cV and and eW, are polarised in the plane of reflexion 

 Sabcde. If we now remove the lower plate D, so as to per- 

 mit the direct pencil a T to reach the eye, the pencils c V, and 

 e W will be mingled with the pencil aT. When RS is inci- 

 dent at an angle of 88° 38', the pencil a T ought to be com- 

 pletely polarised, and ought to vanish in every quadrant, when 

 examined with a polarising crystal. It is prevented, however, 

 from vanishing, by the admixture of the pencils c V and e W, 

 which are polarised in an opposite manner, and which will 

 therefore remain visible when the whole of the direct pencil 

 has vanished. As the angle of incidence diminishes, the in- 

 tensity of the pencils cV,eW suffers a very rapid diminution, 

 while that of the direct pencil aT receives a corresponding 

 augmentation. On this account, the oppositely polarised rays 

 c V, e W form but a very small proportion of the compound 

 pencil, and they are scarcely perceptible when the number of 

 plates exceeds eight. The union of the reflected light with the 

 obliquely refracted pencil, constitutes, therefore, the true cause 

 of the anomalv which we have been considering. 



III. 



