DISCUSSION OF CHEMICAL COMPOSITION. 251 
the biotite departs considerably from its real composition. Biotites 
which are so rich in iron are notoriously variable in make-up. As it 
stands, the calculation indicates a rock composed of 33 per cent biotite, 
60 per cent feldspar, 4 per cent apatite, and half a per cent of magnetite, 
and an inspection of the slides and the amounts obtained on separation 
by Thoulet solution closely corroborate this. 
The presence of chlorine, the apparent excess of alkalies, and the 
close agreement in composition with some nepheline-syenites, which 
will be reverted to later. gave rise to the suspicion, which could not be 
confirmed, that sodalite and perhaps nepheline were present. No trace 
of gelatinous silica could be discovered either in treatment of the slides 
or of the finely powdered rock. Furthermore, careful qualitative tests 
of the apatite, after separation from all other ingredients of the rock, 
except a little biotite, which was entirely unaffected by the nitric acid 
employed, gave such strong reactions for chlorine as to justify the as- 
sumption that all or the larger part of the chlorine present was in that 
mineral, the fluorine shown in the analysis being probably present in 
the biotite. 
The feldspar in this dike must be more largely constituted of albite 
than is the case in the other dikes. While part of the potash calculated 
in the biotite is no doubt replaced by soda, such part is not large in most 
biotites, and even atthe best the amount of soda going into the feldspar 
must far exceed the potash. The testimony of the slides is not especially 
adverse to this. Considerable microperthite is present, but the larger 
part of the feldspar is not microperthitic ; neither is it twinned. This 
untwinned feldspar, on separation, falls at 2.62, the specific gravity of 
albite, the microperthite being a little lighter, so that a fairly good sepa- 
ration of the two may be effected. 
Prrrotogic RELATIONSHIPS 
The magma which supplied the material forming these dikes was 
characterized by high percentages of alkalies, neither predominating, 
by very low lime and magnesia, rather low alumina, and proportionally 
high silica. The mineralogic peculiarities of the rock, paucity of ferro- 
magnesian silicates, prominence of feldspar containing both alkalies, and 
quartz from the silica excess, depend directly on this chemical com- 
position, and the rock has well marked chemic and mineralogic char- 
acters, which permit of no doubt as to its position. The group to which 
it belongs is intermediate between the syenite-trachyte and the eleolite- 
syenite-phonolite groups. With increasing lime and magnesia and de- 
creasing alkalies comes a passage into the former with increasing alumina 
