26 BOTANICAL GAZETTE [JULY 
from youth to age, while in others there may be an increase to a 
maximum followed by decline to the adult intensity, or a fall to a 
minimum succeeded by a return to a rate slightly lower than that 
at the beginning. Study of the filament and anther separately 
reveals the fact that their respiratory intensities are distinctly 
different, the anther undergoing a sort of grand period of respira- 
tory intensity, while the intensity of the filament increases regu- 
larly from immaturity to maturity. The intensity of respiration 
of the stamen as a whole therefore is the resultant between these 
two respiratory intensities. 
Great diversity of opinion exists as to the importance attach- 
ing to the respiratory quotient as an index of metabolism. In 
seeds like Amaranthus and Chenopodium the quotient would 
appear to be of significance because of its stability. The varia- 
bility of the quotient in Rosaceae at first might appear to militate 
against its possessing any significance. When, however, this 
variability of the quotient is found to characterize a group pos- 
sessing fundamental physiological and chemical features in com- 
mon, it would seem that even here some significance might attach 
to the quotient. It may be of little value as indicating the material 
oxidized, but it may have considerable importance as indicating a 
situation due to the interplay of several factors. The quotient 
percentage curves (fig. 4) and the frequency histograms (fig. 3) 
show more clearly than do the tabulated data the general trend of 
respiration. From them can be seen that even with their varia-— 
bility the values for hawthorn and the other Rosaceae tend to fall 
into small groups about one largest assemblage. The latter, there- 
fore, may be considered indicative of the type respiratory value for 
the seed. 
In Chenopodium and Amaranthus the massing of the quotients 
is within a narrow range, and the type is more marked. Treat- 
ment of data in this way, therefore, may serve as a further check 
on the uniformity of conditions under which the experiments are 
carried out, and perhaps on the reliability of the method. 
In this connection it is interesting to note that the curves in 
fig. 3 are of the kind found by Prarson to be typical for botanical 
measurements, limited skew curves (‘‘axis of the abcissas limited 
