n!

108

FLUIDITY AND PLASTICITY

slope. It therefore occurred to them to compare their substances
at temperatures of equal slope and they seemed to find theoret-
ical justification in this proceeding, since at a given slope the
temperature is exercising the same effect upon the viscosity of
different substances, i.e., dy/dt is constant.
They were able to establish the most nearly quantitative
relationship in the comparison between molecular viscosity work
and chemical composition and constitution using a constant
slope, arbitrarily selected as 0.000,032,3. We shall now examine
the nature of this relationship.
By comparing the values for the homologues given in Table
XXII they observed that the addition of a methylene group
to a compound increases the observed value of the molecular
viscosity work by (80 ± 5) X 10~3 c.g.s. units. They assume
that CH2 = 80, the factor 10~3 being understood. Similarly an
iso-grouping is found to lower the value observed for the normal
compound by 8 ± 3 provided that the highly associated butyric
acids are left out of the calculation.
The value of H2 was found by subtracting the value of nCH2,
as calculated from the above constants, from the observed
values of the paraffins whose general formula is CnH2w + 2 as
shown in Table XXIII. The mean value of H2 is —68 and since
CH2 = 80, C - 148.
Comparing normal propyl with allyl compounds, it was
found that the occurrence of a double linkage and the loss of two
hydrogen atoms lower the molecular viscosity work by 27 ± 1;
hence the value of a double linkage was assumed to be —95.
Using the values thus obtained, they determined the value
of oxygen in ketones to be —19, excluding acetic aldehyde and
dimethyl ketone from the calculation because they are the first
members of their respective series, and are probably associated.
In the aliphatic acids the two oxygen atoms have a value of
81 ± 4, but since one of these is a carbonyl oxygen, the value of
hydroxyl oxygen must be 100. On the other hand, oxygen when
united as in ether was found to be 43. It seemed to them possible
that oxygen might have yet other values such as the carbonyl
oxygen in aldehydes as distinguished from ketones. Comment-
ing on the different values which it seemed necessary to give to
the same atom in differently constituted compounds, Thorpe