180 



rate them in that direction. These forces may possibly modify the 

 doubly-refracting structure, but the author has not examined this 

 question. 



On trying certain soft solids which possess double refraction, such 

 as bees' wax, oil of mace, almond soap, and tallow, remarkable results 

 were obtained. Almond soap, the particles of which are not in op- 

 tical contact, may be drawn out into strings, and these strings possess 

 neutral and depolarizing axes like the streaks above described. This 

 is done by traction alone. Similar results are obtained in oil of mace 

 and tallow, by compression and traction. In bees' wax. the depo- 

 larizing lines are even better displayed, especially if a little common 

 resin be added. 



It is not easy to explain why, in these experiments, the optical 

 and crystallographic axes of the particles are placed in the same line. 

 Mechanical force is the primary agent, but it is possible that elec- 

 tricity may also contribute, even in the case of almond soap, to the 

 result. In that case, however, by drawing it out into a thread, we 

 diminish all the lateral obstacles to a crystalline arrangement. Ele- 

 mentary prisms, or crystals whose length much exceeds their breadth, 

 will then tend to place their long axes in the line of traction, and as 

 the lateral obstructions are removed, the particles may follow their 

 natural tendency. 



We have reason to suppose, that in hard substances the same 

 principle acts, and that the particles, when drawn into narrow lines 

 and freed from lateral attractions, may more readily assume the crys- 

 talline arrangement which is natural to them, and is the result of 

 certain inherent polarities. 



In some cases, where the crystalline arrangement was imperfectly 

 produced, the author observed a tendency in the particles to quit their 

 position, as if they were in a state of unnatural tension or restraint. 

 This probably depends on the non-homologous sides of the elementary 

 particles having been brought into contact, a condition quite com- 

 patible with the existence of neutral and depolarizing axes, provided 

 the non-homologous sides deviate from their proper position either 

 90° or 180°. In that case, polarized light, directly transmitted, 

 will exhibit the same colours as if the sides were in the normal po- 

 sition. But if transmitted obliquely, the hemitropism of the com- 

 bination, as we may call it, will be at once detected by the difference 

 in colour of the two plates. 



