Feb. a, 1974 
Tissue Fluids in Cotton 
291 
The correlations between members of the same duplet or triplet 
(column 7) are in all cases higher in the first than in the second series. 
Since the problem of the magnitudes of the correlation between the 
members of the same duplet or triplet is not a subject for special dis¬ 
cussion in the present analysis of the measurements, the probable errors 
of the differences between these correlation coefficients for the first and 
second series of determination have not been determined. 
Finally, we note that the differences between Egyptian and Upland 
cotton are uniformly higher in the second than in the first series of col¬ 
lections. This is clearly shown by the absolute differences in column 8. 
The increase in the absolute difference between the Egyptian and Upland 
cotton from the first to the second series of determinations may be shown 
by taking the differences between these differences. These are shown 
immediately beneath the two differences. 
The increase in the difference between Egyptian and Upland from 
the first to the second series of determinations is in all cases over three 
times as large as its probable error. In three cases it is over four times 
as large as its probable error. 
Correlated with this larger difference we find a uniformly higher 
criterion of significance of difference (column 9) between the Egyptian 
and the Upland types in the second series. 
The constants just discussed represent the absolute differences between 
the Egyptian and the Upland cottons as studied in the first and second 
series of determinations. It must be remembered, however, that in both 
Egyptian and Upland types the osmotic concentration is uniformly 
lower in the second series than in the first series of determinations. This 
means that the differences which have been shown to be absolutely 
larger in the second series than in the first are relatively much larger. 
This is clearly shown by the entries in the column (10) of Table X giving 
the percentage differences between the two types. In the first series of 
determinations, the relative differences range from less than 1 per cent 
to over 3.7 pei* cent. In the second series the percentage differences 
range from 4.7 per cent to 8.3 per cent. 
The significance of these various comparisons will be more fully dis¬ 
cussed in a subsequent section after the values of the other physical 
constants have been taken into account. 
COMPARISON ON THE BASIS OF SPECIFIC ELECTRICAL CONDUCTIVITY 
The statistical constants for the specific electrical conductivity of the 
tissue fluids of Egyptian and Upland cottons, expressed in terms of 
reciprocal ohms, kX io 6 , are compared in Table XI, which is made up 
in the same manner as Table X. 
The results show that in every comparison (column 8) the tissue 
fluids of the Egyptian cotton have a higher specific electrical conduc¬ 
tivity than those of the Upland varieties with which they are compared. 
A comparison of the differences with their probable errors (which have 
been calculated with due regard to the correlation obtaining between 
the constants of plants of the same duplet or triplet), shows (column 9) 
that in the first series all of the differences are over three times as large 
as their probable errors. In the second series all of the differences in 
electrical conductivity are over nine times as large as their probable 
errors. 
