366 POLARISATION. 



interpreted in the same manner as with the cylinder. We will 

 only remark that no object has hitherto been found which comes 

 under this category. 



V. 



ON SOME STEUCTUKAL PECULTAEITIES OF OEGANISED 



SUBSTANCES. 



HITHERTO we have strictly adhered to the supposition that the 

 position of the ellipses of elasticity corresponds, in all points of 

 the substance, to the effect which the superposed elements produce. 

 We have assumed that the double-refracting elements, of which 

 the layers of a membrane a cylindrical or spherical object 

 consist, always exhibit the same regular arrangement, and, based 

 upon this hypothesis, that one axis of elasticity is always situated 

 radially, and the other two tangentially. We have then determined 

 the position of the latter according to the effect which the surface 

 views produce, or left them undetermined if these views acted 

 neutrally. 



This course would be self-evident with crystalline media with 

 suitable arrangement of the smallest particles; with organised 

 media it is always plausible ; but it should be emphatically stated 

 that no absolute necessity exists for the assumption of so regular 

 a structure. Organised substances act in all essential points 

 differently from non-organised ones ; their optical character is not, 

 as in the latter, dependent upon the changes of distance which 

 the smallest particles undergo by pressure or tension, or even by 

 swelling ; it remains constant, even when the changes amount to 

 a multiple of the original distances. We can stretch or bend a hair, 

 a bast-fibre, &c., at will, without altering the character of its colours ; 

 whilst, for example, a fine glass tube, even on very slight curvature, 

 produces the colour which corresponds to the change of distance 

 of its atoms thereby occasioned. Just as little does the swelling 

 of a piece of membrane cause an essential change of the optical 

 properties when it is soaked in sulphuric acid or ammonio-oxide 

 of copper, whereby possibly the thickness is increased fivefold, 

 though the length and breadth but slightly. Hence it follows, 



