On Material Molecules and the Etkerial Medium. 253 



ditions differ under which the crystals are formed. In con- 

 clusion, there appears to he no established exception to the 

 law, that the arrangement of the material molecules in space 

 is definite for a substance of definite composition, and that, 

 therefore, any change in chemical composition must be 

 attended by a change in the arrangement of the molecules, 

 and also in their mass, and hence by a change in optical 

 properties. 



It may be objected that the variations in the angle be- 

 tween the optic axes, observed by De Senarmont, are too 

 large to be accounted for by the small changes in the ar- 

 rangement of the molecules, which could be produced by 

 isomorphous replacement ; but, a priori, there is nothing to 

 show that very small changes in position of the molecules, 

 especially if the forces exercised by the molecules on the 

 etherial medium are sensible only at insensible distances, 

 as without doubt they are, may not produce a change in 

 the magnitude and direction of the resultant action on the 

 elements of the ethereal medium, by no means insensible in 

 its effect on the indefinitely small vibrations of the ether. 

 That a minute change in the position of the molecules does 

 produce a change in optical properties, is shown by the 

 effect of pressure in producing double refraction in jelly, 

 fluor-spar, &c. 



But if the optical properties of crystals are really caused 

 by an action between the material molecules and the etherial 

 medium, then the special optical properties of a given crystal 

 should depend on the special arrangement of the molecules 

 in that crystal, as shown by the form of the crystal, and 

 expressed crystallographically by the system to which it 

 belongs, and by the angles between its faces, that is, by 

 the magnitude of its angular elements. In other words, 

 some connection ought to exist between the form and optical 

 properties of crystals ; and it ought to be possible to express 

 the latter in terms of the angular elements of the crystal, 

 and certain other constants dependent on the constitution 

 of the ether, &c. That such a connection does exist, few 

 physicists would, I think, be found to deny. De Senarmont, 

 however, drew from his experiments on the variation of the 



