I. 0. Chacko—On Cordierite. 463 
organic impurities. Buta sedimentary origin is out of the question 
in the present case. It is held by some that pleochroic halos are due 
to the presence of radio-active substances. I kept an Ilford Empress 
photographic plate exposed in a dark box to a crushed sample of the 
mineral for more than twelve hours, but on development the plate 
did not show any indication of having been affected by the mineral. 
Lamellar twinning is very common. But for its pleochroism the 
mineral may be mistaken for plagioclase on account of the similarity 
of its twinning and refractive index to those of felspar. 
Cordierite is usually negative in optical character. But I think at 
least some of the crystals are optically positive in the present case. 
The optical character of a mineral is not an absolutely invariable 
property of it, especially when the axial angle lies in the neigh- 
bourhood of 90°. When the angle is nearly 90°, a slight variation 
in it may change the character of a mineral from negative to positive 
or from positive to negative. Slight variations, and in some cases 
large variations, in the axial angles of minerals are very common. 
In the case of cordierite the axial angle is very nearly 90°, and 
_ a change in optical character may be expected. At the same time 
it must be remembered that the very fact that the angle is nearly 
90° makes it difficult to distinguish the acute bisectrix from the 
obtuse with an ordinary petrological microscope, and on this account 
there is possibility of error in the observations. Minerals of double 
optical character are not, however, unknown. Among biaxial 
minerals, anthophyllite, penninite, and kammererite are examples. 
Owing to the presence of inclusions and the intimate association of 
the mineral with other minerals, it is not easy to determine its 
density with accuracy. Ina series of separations of a finely crushed 
sample with Klein’s solution, particles of the violet mineral were 
found to sink at various stages of the density of the solution between 
3°00 and 2°60 approximately. The denser particles are regarded as 
having inclusions or as associated with other heavy minerals. For 
the analysis given below such particles of the violet mineral as floated 
in a heavy solution in which a quartz crystal would just sink were 
selected. An analysis of such selected grains of the cordierite gave 
the following results :— 
Water. é : ‘ 4 4 : 1-74 
Silica j 4 : , H M . 49-74 
Ferric oxide . ; x 4 ‘ : 5-65 
Ferrous oxide . ‘ K 4 - : 3-00 
Alumina . ; 4 : i J uno 
Lime , 5 ; ‘ : ‘ 3 1-05 
Magnesia : : : : : 4 4-30 
100-69 
What strikes one first in this analysis is the high percentage of 
ferric oxide. Of the eleven analyses given by Dana of samples 
of cordierite from various localities, nine do not give any ferric oxide 
at all, while two give only 0°63 and 0°38 per cent of ferric oxide 
respectively. On the other hand, the percentage of ferrous oxide is 
low in the present analysis. Magnesium replacing ferrous iron in 
