288 
ELEMENTARY CHEMICAL MICROSCOPY 
ing, cigar-shaped plug. It should be rolled dry and rolled long 
enough to make it fine and even. If the paper is cut, not torn, 
there will be a seam in it, and it cannot be so readily made tight. 
The plug thus formed is inserted in the small end of the tube 
from the outside and worked in by rotating the tube until from 4 
to 8 millimeters of the paper are within. The rest of the paper is 
then cut off a millimeter or two from the end of the tube.” 
The filter is first moistened with distilled water and then in¬ 
serted in the drop to be filtered, suction is applied to the larger 
end and the clear liquid drawn up through the filter into the tube, 
from which it is removed by a capillary pipette or by carefully 
removing the filter paper with a pair of fine forceps and expelling 
the liquid in exactly the same manner as in the Behrens method. 
A tightly rolled cigar-shaped plug of filter paper or fibrous 
asbestos may be inserted in a straight Behrens tube in a similar 
manner to that described above, and will be found to yield even 
more satisfactory results than the fragile drawn-out tube of 
Savage. 
The author has found in certain instances that alundum filters 
have proved of great value. Such filters are made by grinding 
tiny conical plugs from pieces of broken alundum crucibles and 
fusing these plugs into the ends of glass tubes 2 to 2.5 millimeters 
in diameter and 50 to 60 millimeters long. After fusing, excep¬ 
tional care must be taken in cooling and annealing. In like man¬ 
ner porous porcelain plugs may be used, but in such an event 
a powerful suction pump is required, suction by means of the 
mouth being insufficient to cause the passage of the liquid. 
Sublimation. — This operation, though of somewhat limited 
application and comparatively seldom employed in inorganic 
qualitative analysis, is so very important, and of such inestimable 
value in the examination of organic compounds, that every worker 
should become thoroughly familiar with it, particularly with the 
method of performing fractional sublimations. 
The usual method is that of sublimation from one slide to 
another. The material to be tested is placed at the corner of a 
thin slide. If it is a solid it is wise to moisten it with water and 
then dry it thoroughly; this will generally effectually prevent 
