METHODS OF TESTING SLATE. 49 
selected, so as to fairly represent the general quality of the bed. It should be fresh, 
unfrozen, and a half inch thick across the cleavage. At least two section- should be 
prepared — although the more the better — one parallel to the cleavage and another at 
right angles to it, never diagonal to it. The sections should be exceedingly thin, 
much more so than ordinary sections of eruptive rocks, and the slide cover should be of 
the very thinnest kind and should be fastened on with a minimum of balsam to admit 
the use of the highest objectives. Both slides should be examined first in ordinary 
light then in polarized light, with powers ranging from 140 to 700 and in some cases 
even 1,100 diameters. The transverse section will show the quality of the cleavage, 
the false cleavage, if any, and under polarized lisrht whether the specimen is a mica 
slate or a clay slate by the entire matrix becoming in a mica slate four times dark 
and four times light in complete rotation. The observer should, however, not be 
misled by the obscuration of aggregate polarization caused by the abundance of car- 
bonaceous matter or carbonate or hematite. The carbonate, which extinguishes 
irregularly, is very troublesome. This could be eliminated by the application of 
HC1 to the bare section and then recovering it. The aggregate polarization may also 
be obscured by the thickness of the section and the whole analysis thus vitiated. 
Sections parallel to the cleavage reveal the amount of carbonate and indicate the 
probable amount of discoloration by exposure. Both sections, under incident light, 
will show pyrite, if any exists. 
Chemical analysis. — This, in order to give a correct idea of the composition of the 
slate, should not be partial, but complete." Such an analysis should then be com- 
pared with complete analyses of the best slates of like color, and before a final con- 
clusion is reached as to the value of the slate its microscopic analysis and the results 
of the tests of its strength, elasticity, porosity, and corrodibility should be considered 
in connection with its chemical analysis. Merriman concludes from six different 
kinds of tests applied to each of 24 specimens of Old Bangor and Albion (Pennsylva- 
nia) slates, as well as from the results of several general chemical analyses, that — 
The strongest slate stands highest in weathering qualities, so that a flexural test affords an excel 
lent index of all its properties, particularly if the ultimate deflection and the manner of rupture be 
noted. The strongest and best slate has the highest percentage of silicates of iron and alumina, but 
is not necessarily the lowest in carbonates of lime and magnesia. Chemical analyses give only 
imperfect conclusions regarding the weathering qualities of slates and do not satisfactorily explain 
their physical properties, b 
As the discoloration in slate has been shown to be probably due to the oxidation 
of the ferrous carbonate (FeC0 3 ), one of the chief objects of a chemical analysis 
should be to show its amount or absence, but this is just what a chemical analysis 
can not do with so complex a rock. 
Hutchins finds that the presence of chlorite minerals can be detected by heating 
the slate to dull redness, thus dehydrating and discoloring those minerals, then pre- 
paring a thin section of the slate so treated and comparing it with sections of the 
normal rock. c 
The most decisive of all these tests are probably those for strength and toughness, 
as applied by Merriman, and the microscopic analysis. 
It should be added that shearing, compression, and expansion tests of slate have 
been made at the United States Arsenal at Watertown, Mass. 
aSee, on the advantage of complete analvsis, Principles and methods of analysis applied to silicate 
rocks, by W. F. Hillebrand, Bull. U. 8. Geo'l. Survey No. 148; Analyses of rocks and analytical meth- 
ods, Clarke and Hillebrand, pp. 1-64, 1897. 
b Merriman. See Bibliography, p. 144. 
<■ Hutchins, Clays, shales, and slates. See Bibliography, p. 140. 
