f the wave which gives a limit of 0.5 mu. or
1/125,000 of an inch. By using waves of shorter length, as the
ultra-violet light, objects of 0.1 mu. or 1/250000 of an inch can be
seen; but as these methods depend upon photography for the
demonstration of the object the study is difficult. The presence of
objects still smaller than 0.1 m. can be detected in a fluid by the
use of the dark field illumination and the ultra-microscope, the
principle of which is the direction of a powerful oblique ray of light
into the field of the microscope. The objects are not visible as such,
but the dispersion of the light by their presence is seen.

The demonstration that infectious diseases were produced by organisms
so small as to be beyond demonstration with the best microscopes was
made possible by showing, that some fluid from a diseased animal was
infectious; and capable of producing the disease when inoculated into
a susceptible animal. The fluid was then filtered through porcelain
filters which were known to hold back all objects of the size of the
smallest bacteria and the disease produced by inoculating with the
clear filtrate. There are a number of such filters of different
degrees of porosity manufactured, and they are often used to procure
pure water for drinking, for which use they are more or less,
generally however, less efficacious. The filter has the form of a
hollow cylinder and the liquid to be filtered is forced through it
under pressure. For domestic use the filter is attached by its open
end to the water tap and the pressure from the mains forces the water
through it. In laboratory uses, denser filters of smaller diameters
are used, and the filter is surrounded by the fluid to be tested. The
open end of the filter passes into a vessel from which the air is
exhausted and filtration takes place from without inward. The test of
the effectiveness of the filter is made by adding to the filtering
fluid some very minute and easily recognizable bacteria and testing
the filtrate for their presence. These filters have been studied
microscopically by grinding very thin sections and measuring the
diameter of the spaces in the material. These are very numerous, and
from 1/25000 to 1/1000 of an inch in diameter, spaces which would
allow bacteria to pass through, but they are held back by the very
fine openings between the spaces and by the tortuosity of the
intercommunications. When the coarser of such filters have been long
in