MR. TALBOT ON THE OPTICAL PHENOMENA OF CRYSTALS. 
31 
§4. 
I have remarked that the circular crystallization of boracic acid is frequently en- 
tirely superseded by other modes of crystalline formation ; which circumstance ap- 
pears to be chiefly owing to the presence or absence of combined water. Some of 
these variations deserve to be particularly specified. 
1. Instead of circles there often occurs a formation of crystals resembling two op- 
posite sectors of a circle combined. This form may be traced in different crystals, 
from its commencement when the angle of the sector is small, through all degrees of 
increase, until at length the opposite sectors unite and form a complete circle. 
2. The crystals are frequently of a very irregular elongated shape, which does not 
approximate either to a prismatic or a cylindrical form. This stem, as it may be 
termed, subdivides itself at both extremities into an immense multitude of diverging 
fibres, giving it the appearance of a bundle of elastic filaments firmly held together 
in the central part, but with its extremities left at liberty to diverge*. 
3. Another variety resembles in the same way irregular stems or branches, which, 
however, instead of being subdivided, are abruptly truncated at both extremities per- 
pendicularly to the general line of their direction. 
4. Sometimes, on the contrary, the ramification is much more developed, and then 
resembles two plumes united by a common stem. 
5. Crystals of regular geometric form. These appear to require the presence of 
combined water. 
Whichever of these formations occurs, it is for the most part seen in all the 
crystals at once, to the exclusion of any of the other forms. 
§5. 
These crystals generally undergo a spontaneous change in the course of one or two 
days after they have been formed. Those (No. 4) resembling plumes usually break 
up and resolve themselves into small rhombs and other geometric forms. The elon- 
gated crystals (Nos. 2, 3) undergo a remarkable change. They become traversed 
with innumerable fissures transverse to their length, and thus break up into thin 
plates, which either cohere loosely or separate entirely. 
§ 6 . 
All these forms are very pleasing objects for the polarizing microscope. This 
arises from the very high depolarizing power of boracic acid, which enables its 
thinnest plates to exhibit colours of great variety and brilliancy, and causes even its 
dust or smallest particles to appear luminous. The more energetically any substance 
acts upon polarized light, the closer and more crowded are the bands and lines of 
* This appearance is not very uncommon in the crystallization of other substances, though I believe it has 
not yet been described. The divergence of the filaments suggests the idea of electrical repulsion as being at 
least its primary cause. 
