40 
MUSEUM BULLETIN NO. 30 . 
Under this assumption, and leaving out of consideration the 
minor components of the solutions, such as K 2 0, MnO, TiO s , 
and P 2 0 6 , which altogether make up less than 1 per cent of 
any of the rocks, it is seen that the hornblendite alteration 
resulted in slight increase of silica, large increase of iron, mag- 
nesia, soda, and water, with loss of lime and alumina. In the 
aplitic alteration a somewhat larger increase in silica and soda 
took place, whereas water was slightly lost, alumina more so, 
and lime, iron, and magnesia almost completely removed. It 
may also be seen from Figure 1 that in the main these conclusions 
are correct, whether the hypothesis of constancy of volume be 
accepted or not, as most points are so far to the right or left 
of the assumed zero vertical that, in order to alter their signifi- 
cance materially, quite inadmissible assumptions as to weight and 
volume changes would have to be made. Thus, for instance, to 
assume that magnesia had not been added during the hornblend- 
izing alteration, would be equivalent td assuming that the horn- 
blendite possesses approximately 30 per cent of pore space, which 
is contradicted by field and microscopic evidence. 
Composition of Solutions. 
The homblendizing solutions reacted with the gabbros very 
readily and rapidly, judging from the wide bands of altered rock 
which have resulted from the percolation of solutions along very 
narrow joint cracks; and in no case was it found that further 
alteration took place after the rock was once thoroughly altered 
to hornblende, although in some cases additional magnetite was 
deposited. It is reasonable to conclude, therefore, that the horn- 
blende was in equilibrium with the solutions when the alteration 
was complete, except that the solutions were apt to be still 
supersaturated with iron. Analysis No. 2, therefore, an example 
of thorough alteration so far as the microscope can show, repre- 
sents the composition of the solid in equilibrium with the liquid 
which produced the alteration. Such an equilibrium implies the 
presence in the solution of all the component oxides indicated in 
the analysis, though not necessarily in the same proportions as in 
the hornblendite; in fact, the solution must have been saturated 
