The President's Address, H. C. Sorhj, P.G.S. 11 
and object-glasses. I find that they can he greatly improved by 
using a condenser of shorter focal length, specially constructed so 
that it can be adjusted for different thickness of glass. The object- 
glasses should also possess special characters. It is more important 
that the spherical aberration should be well corrected, than that 
the aperture should be large, or the image achromatic, since red 
light is generally used for illumination. 
When the thickness of a section is considerable it may be 
allowable to assume that the polished surfaces are in contact with 
the thick glass plate on which it is mounted, and with the thin 
glass cover; and we may disregard the thickness of the intervening 
Canada balsam ; but when we come to deal with very thin sections 
this would lead to very great errors. It is also necessary to bear 
in mind that both the thick and the thin plate glass are not of 
uniform thickness, so that measurements made in any one part 
would apply only to closely adjacent parts. It is also very neces- 
sary to remember that the thickness of any transparent substance 
measured in the manner described, by looking through itself is not 
the same as its real thickness, but is approximately equal to its 
real thickness divided by its index of refraction. In accordance 
with these principles the most legitimate process appeared to be to 
measure the total thickness of the mineral, of the upper and lower 
balsam, and of the covering glass, and to deduct from it the true 
thicknesses of the balsam and of the glass, calculated from their 
apparent thickness ; and also to determine the displacement of the 
focus due to the mineral alone, by similar means. Such a process 
is, however, very tedious, and the chances of error are greatly 
increased by the large number of measurements required, and I 
have therefore been led by degrees to greatly simplify and improve 
it. My experience so far leads me to recommend three different 
methods, one or other of which appears to be the best, according 
to circumstances. If the specimen be somewhat thick, and the 
indices of the minerals to be observed not so different from that of 
Canada balsam as to make a slight error in its thickness of 
importance, the balsam between the glass slide and the covering 
glass should be carefully cleaned out along one edge of the section. 
Fig. 7 shows what would then be the general relation of the 
different parts as seen in section. The thickness and effect of the 
covering glass { g) may then be entirely neglected, since the true 
distance between it and the glass slide is easily measured, and so 
is the total displacement of the focus due to the mineral (m) and 
the upper and lower balsam {h, U). Unless the balsam be rela- 
tively very thin its apparent thickness must be measured and due 
allowance made for it in calculating the results. The mineral 
observed ought to be as near to the edge as possible, to avoid any 
errors due to varying thickness. 
