126 SCIENCE PROGRESS. 



posed to consist of birefringent uniaxial particles arranged 

 with their axes along the meridians ; during its formation, 

 since any elastic force which might act in opposition to the 

 surface tension is lacking, the particles would flow from one 

 part to another in such a way as to make the surface 

 potential energy a minimum, and the crystal would, like a 

 drop of amorphous liquid matter, assume the form of a 

 sphere. In the ordinary solid crystal the surface tension 

 is different for different faces, and the crystal assumes a 

 polyhedral form. The figures which accompany the memoir 

 also illustrate the curious manner in which the drops may 

 be deformed and broken up by squashing them between 

 the slide and a cover glass. A globule may be broken up 

 into two identical globules in this way ; two distinct 

 globules may even be brought into contact and merged 

 together so as to constitute a single globule. 



The question next arises whether the isotropic liquid 

 into which the liquid crystals are converted by further 

 heating is also a crystalline li(|uid belonging to the cubic 

 system, in which case the transformation of the drops to the 

 isotropic liquid is merely a passage of a liquid Irom one 

 crystalline modification to another, analogous to the 

 passage of the birefringent lamellae of boracite to the 

 solid cubic modification when a crystal is heated to 

 265°. Now it is found that when boracite and similar 

 substances are transformed into allotropic modifications the 

 new crystals are regularly orientated with regard to the old. 

 In the same way when the liquid crystals are transformed 

 into the isotropic liquid by rise of temperature, and back 

 again to the birefringent drops by fall of temperature, the 

 latter preserve their original orientation. It is as though 

 the terrestrial globes which they resemble had been fused 

 into a liquid mass, and on separating again from this on 

 cooling, solidify with their axes j)ointing the same way as 

 before. 



Hence, it is argued, the isotropic liquid may be only a 

 new crystalline modification of the liquid crystals, and that, 

 just as is the case in the transformation of solid crystals, 

 the one modification may be converted into the other 



