ON LIGHT. 375 



square E F G H. If the initial difference of phase be 

 two, three, or four twentieths of the period, it will be 

 seen, by following out the movement in the same way, 

 that more and more open ellipses will come to be 

 described as represented in the figure ; and that, when 

 this difference amounts to five twentieths or a quarter 

 undulation, the movement will be circular in the direc- 

 tion p P q Q, or of the arrow marked thus -|-. The 

 difference of phases still continuing to increase, this 

 will again degenerate into an ellipse by a continued 

 elongation in the direction H F, and contraction in 

 the direction E G, till 'it passes at length, after another 

 quarter-undulation of phase-difference, into the straight 

 line H F. The circulation in all, however, being in the 

 same direction, or -f. On the other hand, if instead of 

 supposing the vibration a b to be initially in advance of 

 A B by one-twentieth, we suppose it to be so much in 

 arrear, we shall have the same ellipse n E m described 

 as in the former case, but in the opposite direction, that 

 of the arrow marked , as will be easily seen by going 

 through the successive steps of our reasoning : and so 

 for all the rest ; so that in the case of a circular revolu- 

 tion, the direction of the rotation will be one way or the 

 other, according as the vibration a c is a quarter-undu- 

 lation in advance or in arrear, in respect of phase, of A c. 

 (152.) This, then, is what is meant by circular and 

 elliptic polarization. It is easy to extend the reasoning 

 above stated to cases in which the component vibrations 

 are of unequal intensity (or extent of excursion), and 

 make other than a right-angle with each other's direc- 



