312 Journal of Agricultural Research voi. xxvn, no. 5 
Now the actual results for percentage change in concentration from 
the first to the second series are exactly the opposite of what might be 
expected from the hypothesis and from the apparent facts with regard 
to the fruiting of the cottons. Since the Upland cotton was well into 
the fruiting condition when the first series of determinations was made, 
while the Egyptian cotton was just beginning to form bolls, one might 
on the assumptions made expect to find the greatest change in the sap 
properties of the Egyptian plants since it is in these plants that the 
greatest change in physiological state with respect to fruiting seems to 
have taken place. The results are just the reverse. 
It is interesting in this connection to compare the results for the hybrid 
and the parent forms. The hybrid series show a greater percentage de¬ 
crease in the value of A than either the Upland or the Egyptian plants, 
and there was presumably less change in the condition with respect to 
fruiting between the first and the second series than in the case of the 
Egyptian cotton. 
THE RELATIVE MAGNITUDES OF THE DIFFERENCE BETWEEN THE TWO 
SPECIES IN THE FIRST AND SECOND SERIES OF DETERMINATIONS 
In the discussion of the individual constants we have noted that the 
difference between the Egyptian and Upland type is generally larger 
in the second than in the first series. This is true for freezing-point de¬ 
pression, specific electrical conductivity, and generally for the ratio of 
specific electrical conductivity to freezing-point depression. The differ¬ 
ences between the differences in the varieties in the first and second 
series may be considered significant in the case of the two direct meas¬ 
urements, but are not sufficiently large to be regarded as certainly 
trustworthy in the various subseries in the case of the ratio. The results 
for hydrogen-ion concentration are uncertain since the signs of the differ¬ 
ences between the differences are not wholly consistent, and only one 
of the differences (that for rows 3 and 7) may be considered statistically 
significant in comparison with its probable error. 
It is a significant fact that while the second series uniformly shows a 
larger differentiation of the Egyptian and the Upland types for both 
freezing-point depression and specific electrical conductivity, the actual 
magnitudes of the physical constants are lower for both freezing-point 
depression, A, and specific electrical conductivity, k , in the second 
than in the first series. The ratio k/ A is on the average consistently 
lower in the second than in the first series, but the differences between 
the means are small and can not be considered significant individually 
in comparison with their probable errors. This is also true for the 
hydrogen-ion concentration. 
For convenience of comparison, all of the percentage differences be¬ 
tween Egytian and Upland cotton given in column 10 of tables X to XIII 
have been laid side by side in Table XX. 
