50 Dr. Brewster’s experiments on the 
When a body is composed of two or more films whose 
neutral axes are nearly coincident, the compound film will 
exhibit the fourth kind of depolarisation, and the approxima- 
tion to a neutral axis will be more or less perfect, as the co- 
incidence of the axes is more or less complete. This pheno- 
menon is most likely to be exhibited by thin plates which are 
composed of a small number of films. 
Hence every body, in which there is an approach to a neu- 
tral axis, must be composed of two or more films, whose 
neutral axes happen to be nearly coincident. These films are 
probably deposited and crystallized at different times ; or if 
their crystallization has been simultaneous, the forces, or 
causes by which it was produced, must have acted indepen- 
dently of each other. 
5. The fifth kind of depolarisation takes place when there 
is an approach to a depolarising axis, or when the crystal 
restores only a portion of the vanished image. 
If we suppose one part of a body to have no crystalline 
texture, while another part of it has the structure necessary 
to depolarise light, it will exhibit correctly the fifth kind of 
depolarisation. The uncrystallized portion being incapable 
of restoring any part of the vanished image, the light which 
it transmits will form no part of the depolarised pencil, which 
will consist merely of the rays transmitted through the struc- 
ture which has the property of double polarisation. The mag- 
nitude, therefore, of the depolarised pencil will be a measure 
of the portion of the substance which has undergone crystal- 
lization. From the phenomena of caoutchouc, this explanation 
derives great support. When the crystalline texture of this 
substance has been destroyed by heat, it ceases to act upon 
