Anniversary Address of the President, H. C. Sorby, F. B.S. 138 
why so many have been far more worn and rounded than the 
grains met with in subaqueous deposits. Both in them and the 
sand of the Desert, other things being equal, the amount of 
wearing is greater in the case of the larger than in the case of the 
smaller grains, which is easily explained, since their weight and 
the amount of friction would necessarily increase in higher pro- 
portion than their diameters. The contrast between this sand of 
the Desert and equally coarse sand from the Millstone Grit is most 
striking, as will be seen on comparing Figs. 13 and 15, and points 
most clearly to a very different history, although the materiak 
was in both cases originally derived from granite. 
Sand derived from Schists—On the sides of the valley of 
the Tay, north of Dunkeld, occur terraces of sand at some eleva- 
tion above the river. This sand is fine-grained, the common 
size of the grains being about st> of an inch. ‘There is no 
difficulty whatever in seeing that a very large proportion are 
flat plates, either by making them turn round in water, by care- 
fully studying their appearance when the focus is slightly changed, 
or by examining their action on polarized light, when mounted 
in balsam. Fragments of dark green hornblende are common. 
Since we cannot examine sections of the mica in the proper 
direction, its dichroism cannot be observed, but some of the 
darker flakes may have been derived from granitic rocks or 
highly altered schists. On the whole, the microscopical characters 
clearly indicate that the great bulk of the deposit was derived 
from schists; and, considering the geological character of the 
surrounding country, there can be little doubt about the accuracy 
of this conclusion. 
In the neighbourhood of Moffat and of Bangor occur hard slate 
rocks without cleavage, which can easily be cut into thin sections 
perpendicular to stratification. These show alternating layers of 
coarser and finer grain, and of red or pale green colour. The 
erains of quartz sand in the coarser layers are seldom y}5 of an 
inch in diameter. Those of 335 to s}y are common, and many 
are yoyo or less. A large proportion of these are flat plates, like 
those derived from schists, and in some cases they enclose plates 
of mica parallel to their longer axis, as shown by Fig. 16, 
the shaded part being the mica. Such a grain may be looked 
upon as a fragment of mica schist. The quartz is usually 
broken up into such small fragments that only a few show 
complex structure. Well-marked grains of green hornblende 
occur, and a good deal of green dichroic mica. ‘There are, 
however, a few which are dark dichroic, and in the very fine- 
grained layers are many minute granules, having all the characters 
of those derived from felspar or other minerals of granite rocks 
