326 Dr. J. Kerr's Experiments on the 



Proposition II. 

 Of the two pencils in the foregoing proposition, the one which 

 has its plane of polarization parallel to the line of strain is the 

 more sensitive, being most retarded or most accelerated of the 

 two by any given compression or tension. 



5. The proof is simplified by a little change of method. The 

 slit-screen is removed from the plate, and a narrower screen 

 is substituted, covering the neutral band and leaving more 

 than half an inch of glass margin on each side. The experi- 

 ments may be taken in four cases as formerly, and in the 

 same order. 



(1) and (2). Both pencils through one margin, V being 

 outside, that is, in the place of greater strain. The effects 

 are not clearly distinguishable in any respect from those in 

 (1) and (2) of the last article, the pencil V being relatively 

 retarded by the strain in the compressed margin, and rela- 

 tively accelerated in the extended margin. 



(3) and (4). Both pencils through one margin, H being 

 outside, or in the place of greater strain. The pencil H is 

 relatively and faintly retarded by the strain in the compressed 

 margin, relatively and faintly accelerated in the extended 

 margin. The effects are of the same kind as those in (3) 

 and (4) of the last article, but fainter in a striking degree 

 (for any given strain) than those in (1) and (2). I find, 

 indeed, that the effects in the present cases (3) and (4) are 

 sometimes barely or not at all perceptible, especially when 

 the plate and margins are very wide, and this even with plates 

 which are very thick and strained nearly up to fracture. As 

 the pencils pass now through one margin, they are both 

 accelerated or both retarded ; and it follows evidently that 

 the pencil V, which is in the place of weaker strain, is the 

 more sensitive. 



Like the former experiments, these improve on repetition, 

 the clear contrast between the two pairs of cases coming out 

 with perfect regularity. 



Proposition III. 

 To explain the action of directionally strained glass in the 

 common polariscope ; or to infer Brewster *s property of strained 

 media from facts more fundamental. 



6. Brewster's property may be described thus. When the 

 strained plate of proposition I. is placed between a pair of 

 crossed Nicols, it acts as a birefringent medium ; and if the 

 plate of glass be superposed upon, or merely compared with, 

 a plate of uniaxal crystal with optic axis parallel to line of 

 strain, the same component ray is relatively retarded in both 



