Gabbroid Rocks of Minnesota. — Winchell. 175 
the pyroxene is filled with very fine particles similar in all re- 
spects to the larger lamellar inclusions, whereas the area im- 
mediately about the larger inclusions contains no such par- 
ticles. Such a condition is clearly not incompatible with the 
idea of a primary origin for the inclusions ; indeed every thin 
section of holocrystalline rocks shows analogous phenomena ; 
it will only be necessary to mention the way in which apatite 
crystals grow to a size out of all proportion to the amount of 
phosphoric acid contained in a rock. This is an extreme case ; 
most minerals possess the power of attracting to themselves, 
when crystallizing in a magma, the elements necessary to con- 
tinue their growth in a much less marked degree. And in the 
case before us the power of attraction is evidently very weak 
since it is only exercised through such small distances as one 
or possibly two hundredths of a millimeter. 
Therefore we must conclude that there is no good evidence 
showing that these inclusions are not primary, while the fact 
that inclusions of very similar nature are sometimes 
of primary origin, as elsewhere shown*, lends probab- 
ility to the belief that these have the same origin. This con- 
clusion is supported by the fact that the inclusions are very 
rarely limited by planes of the surrounding pyroxene, -but are 
usually wholly irregular in outline, though they also occur with 
outlines (e. g. hexagonal) manifestly peculiar to themselves. 
Olivine alters so universally and in so many different ways 
that it is extremely difficult to distinguish the primary inclu- 
sions from the alteration products. The former seem to be de- 
cidedly rare ;' however liquid inclusions occur very uncom- 
monly, and certain magnetite crystals have presented the ap- 
pearance of primary inclusions ; although, of course, much the 
greater part of the included magnetite is undoubtedly second- 
ary. 
No primary inclusions have been noted in the magnetite. 
The apatite contains occasional grains of a black opaque min- 
eral, doubtless magnetite ; liquid inclusions also occur, though 
not commonly. 
Alterations. The gabbro is in general very slightly altered ; 
it is for this reason that it is possible by its study to obtain 
clear conceptions of the true original character of a newly 
*See Chapter IV. (inclusions in labradorite). 
