124 Transactions of the Royal Microscopical Society. 
facts, even though they are more or less affected by exceptional 
differences. For this reason, in studying the micaceous con- 
stituents, we are compelled to place far more reliance on mere 
colour than would be admissible in accurate mineralogy ; but still 
the difference of colour is often so well marked, that it does appear 
to be characteristic, even although probably the same mineral 
species may be thus separated and different species united together. 
It must also be borne in mind that when I speak of mica I do so in 
a very general sense for micaceous minerals, which may or may not 
be true mica, and may often be more closely allied to chlorite. I 
use it in fact in such a sense as is convenient for the purpose now 
in hand. If we were to be content with nothing short of minera- 
logical accuracy, it would be necessary to abandon the study in 
despair. or this reason it is convenient to classify the micaceous 
constituents as colourless, brown dichroic, and green dichroic. 
The colourless varieties appear to occur both in granites and 
schists, and no reliance can be placed in them to establish any 
difference. On the whole, the dark brown-red or nearly black 
dichroic varieties are characteristic of granites or very highly 
altered schists, whereas the more or less green dichroic are charac- 
teristic of less altered schists and slates. 
Hornblende and Schorl.—The hornblende of schistose rocks 
usually occurs in crystals considerably longer than broad, and 
when broken up gives rise to fragments bounded by parallel sides, 
with more or less rectangular ends, as shown by Fig. 7. 
The colour is usually different shades of green, varying with 
the position of the crystal, and showing blue-green and yellow- 
ereen dichroism. Though sometimes occurring in granitic rocks, 
hornblende is far more abundant as a constituent of schists in 
those districts of England and Scotland which I have more par- 
ticularly examined, and its presence in the associated stratified 
rocks of more recent date must be regarded as indicating that they 
were derived more from the decomposition of schists than of 
granites. 
Schorl not having a well-marked cleavage does not break up 
into such regular shaped fragments as hornblende. The intense 
dichroism of the coloured varieties is a most important character. 
On the whole, it may be looked upon as more characteristic of 
granitic than of schistose rocks, though it is probably in some 
cases more abundant in the latter than is commonly supposed, and 
thus no very definite conclusion can be drawn from the presence of 
detached fragments in sandy deposits. 
Felspars.—Except in a few cases where grains of the original 
felspar sand can still be seen in mica schist, felspars are almost or 
quite absent from this class of rocks. For the purposes of our pre- 
