interval for the difference in means for quartz, feldspar, and biotite was calculated. If this interval included zero, then I concluded that there was enough statistical evidence to reject H0• The results of the statistical work indicated that there is not enough evidence to reject H0 in favor of Ha for quartz and biotite. However, the data also indicate that Ho should be rejected in favor of Ha for feldspar. Though this result is difficult to interpret, it may simply be due to the difficulty of distinguishing untwined feldspar from quartz in unstained thin sections. Though there is clearly a need for more work on the subject of this thesis, the results of my work point to the potential usefulness of point counting scanned images to determine the modal compositions of rock specimens.