53 
of Edinburgh, Session 1872 - 73 . 
defined zones surrounding a well-marked central apex. With 
higher optical powers a greater number of segmentary zones are 
brought out, and radial and transverse stride in exquisite detail. 
Besides these individual characteristics, they exhibit composite 
relations of a peculiar kind, but in reality the development of a 
very simple principle. Whenever an increase of size has produced 
the contact of two or more individuals of a group, further enlarge- 
ment has taken place by the formation of a common investing border. 
From the deposition of the objects in successive layers, with their 
conical extremities resting on what had been at some stage of the 
crystal’s formation one of its faces, they must have obtained their 
position whilst the crystal was in the act of formation. 
Sections, either perpendicular or at an oblique angle to a face of 
the crystal, by presenting a side view of the objects, show that the 
hemispheroidal upper and the conical lower extremities were 
generally connected by a cylindrical body, whose comparative 
length varied considerably in the different individuals. 
The same sections also exhibit a remarkable structural relation 
between the superimposed layers which occupy the successive 
laminae of the crystal. When the objects can be traced from the in- 
terior towards the surface, they are found to have a linear arrange- 
ment symmetrically round axes perpendicular to the respective 
faces. These structural features can be well observed in sections 
through the terminal pyramid of some specimens of amethyst, in 
which they constitute a number of groups equal to the sides of the 
pyramid, each group consisting of a series of highly ornate beaded 
columns perpendicular to the same plane, and therefore parallel to 
one another. 
The individual objects in the same axial line or column are often 
joined so closely that they may be considered as segments of a con- 
tinuous whole, but in other instances the connection is a mere 
microscopic filament. In parts of the crystal a whole series of 
columns terminates on the same lamina, where the last segment of 
each has spread out to an extent which gives the structure the 
appearance of a disc with a long beaded handle attached to its 
centre. This last circumstance indicates one or both of two con- 
ditions — a period of retardation in the increase of the crystal, or of 
rapid acceleration in the growth of the objects. Irrespective of the 
