53 
and this not only in different crystals, but also, to a less extent, in the 
same specimen. As illustrations of the form of such cavities, we refer to 
the Plate figs. 1, 2, 3, and 4, the extent to which they are magnified 
being shown in each case. At the ordinary temperature the bubble in the 
cavity shown by fig. 1 is about one half its diameter, but disappears en- 
tirely at 30 Q C. By carefully measuring the size of the cavity in various 
positions, and comparing it with the diameter of the bubble at 0 Q C. , it 
appears that the liquid expands from 100 to 152 when heated from 0° to 
30° C. Fig. 2 is a tubular cavity, and shows in a very excellent manner 
the boiling of the liquid when it cools, after having been made to expand 
to fill the whole space. At the ordinary temperature the liquid occupies 
only about half the cavity; but when heated in a water-bath to 32° C, it 
fills it entirely. No bubble is formed until the temperature has fallen to 
31° ; and then innumerable small bubbles are suddenly formed, which rise 
to the upper part and unite ; but instead of the liquid merely contracting 
by further cooling, it still continues to boil for some time, as represented 
in the drawing. Two other large cavities contained in the same specimen 
also behave in the same manner, and become full and suddenly boil at 
almost absolutely the same temperature as that figured. We need 
scarcely say that such cavities are extremely rare, and are very remark- 
able, even when looked upon as microscopical objects, independently of 
their interest in connexion with physics. Fig. 3 is a tubular cavity of 
more irregular form, and is interesting on account of there being two 
plates of the sapphire projecting into the cavity, so as to nearly divide it 
into three portions. At the ordinary temperature these partitions prevent 
the passage of the bubble from one part to the other; but by breathing on 
the object through a flexible tube, the slight increase of temperature 
expands the liquid, so as to make the bubble small enough to pass into 
the next compartment ; and a repetition of the process causes it to pass 
into that at the other end. Such plates projecting into the cavities are 
very common; and it is requisite to pay attention to this fact, since 
otherwise they might easily be mistaken for crystals of some other 
substance included in the cavity, which, if they ever occur, must be 
extremely rare, since no decided case has come under our notice. 
In examining sections of sapphire cut in a plane more or less parallel 
to the principal axis of the crystal, the double refraction is so strong, that 
two images of every object lying at any depth below the surface are seen, 
in such a manner as to make them very confused. This may be avoided 
by using polarized light without an analyser, and arranging the plane of 
