VISCOSITY AND FLUIDITY

105

tion is the characteristic additive property, and not these quan-
tities themselves.'7 This conclusion was based upon data which
for the reasons already given was not well suited for reaching a
final -decision of the matter. As the result of more recent study
of the matter with Monroe (1919) and Wright (1920) Kendall has
come to the conclusion that no formula tested by him will repro-
duce the observed data. The present author is in hearty accord
with this conclusion of Kendall. It cannot be emphasized too
strongly, to the novitiate particularly, that no single formula will
reproduce faithfully any considerable portion of the observed
data on the fluidity of mixtures. Moreover it is useless to look
for such a formula in the present state of our knowledge. A much
better plan is to assume the additivity of fluidities, which also
has the virtue of being the simplest hypothesis that we can make,
and then try to account for the deviations from the exact law
on the basis of well-established physical and chemical evidence.
If the fundamental hypothesis is incorrect, incongruities will soon
develop to put us on the right track. If correct, we should pro-
ceed as rapidly as possible to exploit the new knowledge which
fluidity measurements place in our hands.