66

FLUIDITY A.ND PLASTICITY

?j'

I

large. (2) The time of flow should be small for the sake of
economy and also that the temperature may be more readily kept
constant during the time of flow. (3) Small masses of fluid corne
to the temperature of the bath more quickly, and (4) there is
an economy in material, which is sometimes very important.
The minimum of flow is determined by our ability to read the
volume with the desired accuracy. This in turn is determined by
the diameter of the constricted portions of the instrument above
and below the measured volume V. If, however, the constricted
parts are of very small bore, the capillary action becomes dis-
turbing. Very viscous liquids will not drain out properly and
they may even form a meniscus across the capillary which will
prevent the transmission of the pressure and will render the
results quite valueless. It may be remarked that the troubles
due to bad drainage may be minimized by having the drainage
surfaces everywhere as nearly vertical as possible. In other
words, the change from constricted portion to the tube of large
diameter should be made gradually. If the constricted part of
the instrument has an inside diameter of 0.25 em we believe
that the capillary correction will not cause trouble. The
volume per centimeter of the constricted tube is then 0.05 ml
and if we assume that the meniscus can be read to 0.01 cm as it
passes a mark on the tube, it is only necessary to have a volume
of 0.5 ml to obtain the desired accuracy. To provide a margin
of safety in the construction and use of the apparatus we select
about 3 ml as the minimum.
To detect any error due to faulty drainage, it is only necessary
to test the flow of the most viscous liquid to be measured using
very different rates of transpiration by varying the pressure.
Lack of perfect drainage will be made evident, by the substance
appearing to be more viscous at the lower rate of flow. Natur-
ally the more viscous liquids must be allowed to flow slowly
enough so that the drainage will appear to be perfect. If in the
instrument depicted in Fig. 23 the flow were to begin with the
upper meniscus at the point marked 3, it would be necessary for
all of the liquid of the measured volume V to have drained out at
the expiration of the time t. This is not necessary, however, if
the flow begins at some point considerably higher up, as for exam-
ple in the neighborhood of the trap-opening F, for then a certain