LITHOLOGY. 153 
often seen to be hexagonal. Here and there in the mixture of carbon- 
ates a kernel of augite is found; apatite needles are very abundant; also, 
crystals of magnetite. Here and there a well formed crystal of horn- 
blende is seen, and also a bit of pyrite. The chlorite that is very spar- 
ingly present can be called nothing better than viridite, for it is in 
minute, formless bits, which in polarized light behave like an amorphous 
substance. 
As the feldspar is quite fresh and predominant in amount, an analysis 
of this rock may have some value as indicating its original nature. It is 
indeed interesting to note that with the complete destruction of the 
augite, the feldspar has so well maintained its identity. This rock from 
Bemis brook gave Mr. Pease, of the Sheffield Labratory,— 
Silica, ; - . ‘ ‘ . . ‘ ei é : 47 64 
Alumina, . : : ‘i ‘ : - z z : e , 18.35 
Tron sesquioxide, . 5 ‘ : . . z 5 ‘ 2 4.20 
Tron protoxide, 5 . ‘ ‘ a F a 5 A 5 6.52 
Manganese protoxide, . 3 ‘ 2 ‘ F 2 .16 
Lime, 3 2 ‘ 3 a . r F 3 F j : 7.08: 
Magnesia, é é ‘ é . . A i : . ‘ 4.36 
Soda, . . é a ‘ c a : . ‘ : 3-31 
Potash, . - 7 : F : : i : ‘ : 1.96 
Water, . 5 7 - : és 7 : 3 . : 5 2.33 
Carbonic acid, é é ‘ ‘ : ‘ 7 * é ‘i 5.01 
100.92 
The calculations on analyses of such heterogeneous mixtures are not 
very satisfactory, and if we allow at the highest twenty per cent. for car- 
bonates, iron oxide, &c., the silica is increased to sixty per cent., but as 
we know that silica that is liberated by decomposition is often present in 
such rocks in unrecognizable form, the probability that these are labra- 
dorite rocks is indicated. 
The variations in this variety of diabase are not wide. A specimen 
from Tripyramid mountain contains much more unaltered augite, but 
is otherwise the same. The Flume at Lincoln is made by the disinte- 
gration of a diabase identical with that from Bemis brook, save that it 
contains more iron oxide and pyrites, which aid its decomposition. A 
dyke at Dixville is the same, but more green chlorite has been formed, 
VOL. IV. 20 
