86 On a Possible Explanation of the Method employed hy Nohert 
that all the large plates have been cut into slides 3x1 inch, 
in the direction indicated by the observations. In general, the 
direction of the grain can be detected at once by the appearance 
of lines as fine as 30,000 to the inch, while coarse lines may retain 
their initial character for several days. 
The present indications are that the grain is only surface deep, 
and that it is the result of polishing in one direction. Common 
window glass seems to be wholly free from it. Nobert's lines 
are ruled on microscopic cover-glass about ^^-^ of an inch in thick- 
ness. The evidence of grain in this kind of glass is strong ; but 
it is hardly decisive. In some specimens it is very marked, while 
in others it seems to be entirely wanting. Indeed, any conclusions 
on this subject must be regarded as only provisional, owing to the 
extreme difficulty of separating the action of the cutting crystal 
upon the glass from the effect due to the character of the glass 
itself. It is, however, safe to say that in certain kinds of glass 
the best results can only be obtained by ruling in a given 
direction. 
In order to rule bands with lines separated by intervals, e. g. 
of -g^oi^oo- inch, it is of course necessary to rule single lines 
whose width is less than this. Great precaution is requisite here, 
in order to avoid optical delusion. Every microscopist is familiar 
with the phenomenon of false lines. To avoid errors from this 
source, a few single lines are ruled between two heavy finding 
lines. They are then filled with graphite. This precaution is 
necessary in order to give both visibility and distinctness to the 
edges. If the lines are not filled, they may appear much finer 
than they really are ; that is, the objective being in focus for the 
bottom of the furrow may fail to reveal abrasions of the surface 
on either side. The graphite of the New York Graphite Com- 
pany will easily fill the finest line that can be ruled with a 
diamond. 
In order to measure the width of the lines, the following plan 
is adopted as presenting some advantages over the usual method of 
estimating it by comparison with the known value of a given 
division in the eye-piece micrometer. 
First, a single line is ruled, which in the eye-piece apparently 
exactly covers the line to be measured under the objective. A few 
trials will suffice for this purpose. Having found what weight 
must be applied to the diamond to produce such a line, the next 
step is to ascertain how many lines, exactly like this one, can be 
ruled within the space of of an inch with a minimum space 
between each line. This will also require a few trials. For ex- 
ample, if with a yi^th objective and a B ocular, the space -^^^ of an 
inch in the eye-piece corresponds to Y-^^-^-n ^^^^ under the 
objective, and if it is found that fifteen lines can be ruled within 
