and Chemical Composition of Minerals . 5 
lar circumstances, as we have not been able to produce a similar 
salt.” 
In examining several specimens of Sulphate of Potash , I 
found that those which were crystallised in the form of the 
the rhomboidal prism * * and the bipyramidal dodecahedron *f*, 
had two axes of double refraction , while those which crystallised 
in hexaedral prisms had only one positive axis. M. Berzelius 
had the goodness to analyse for me these two different salts. 
The first he found to be the common sulphate of potash ; but 
the second proved to be a double salt, composed of one atom 
of sulphate of potash and one atom of proto-sulphate of iron, 
with water of crystallisation. 
The resemblance between Talc and Mica having induced se- 
veral mineralogists to consider them as the same mineral, I be- 
gan in 1816 to make a collection of different kinds of mica and 
talc, with the view of investigating their optical structure. The 
results of these experiments were sent to Sir Joseph Banks to- 
wards the end of 1817, and published in the Transactio for 
1818. I found that talc was essentially different from the ordi- 
nary mica, the former having its resultant axes inclined at an 
angle of 7° 24' ; while the inclination of the axes in one kind 
of mica was 45°. I found, also, that another species of mica 
had its resultant axes inclined 14°; and that Lepidolite had the 
same optical structure as the Siberian mica J. I sought in vain 
for that kind of mica which M. Biot had found to possess only 
one axis of double refraction ; though I have since foun the 
same property in mica from Karioet in Greenland, and also in 
Mica containing Amianthus. The removal of one of the axes 
from these specimens of mica, was ascribed by M. Biot to imper- 
fect crystallisation , which he supposed to produce an “ infinity 
of axes' 1 ' 1 in the plane of the laminse, arising from “ the crossing 
of the axes of the integrant molecules ||.” I endeavoured in 
vain to understand how such an effect could be produced ; but 
* See Phil. Trans. 1818, p. 211. 222. 
*}• See this Journal , vol. I. p. 6. where it is shewn that the bipyramidal dodeca- 
hedron is in this case a compound form. 
* Phil , Trans . 1818, p. 23. 
|| See Mem. Institut. 1812, p. 316. 334, 335. 346, 347. 351, 352. ; and Traitfe 
de Physique , tom. iv. p. 543, 544. 553, 554, 
