734 
DUNCAN S. JOHNSON 
for over thirty generations without any indication of change of 
the strictly unisexual condition. It must also be remembered 
that such perennial dioecious plants as Cycas, Ginkgo, Populus, 
etc., have been under continuous observation for years, without 
any completely authentic case of change of sex being recorded, 
so far as the writer has been able to learn. It may be considered 
as doubtful whether the apparent microspores found by Cham- 
berlain ('97) on the carpel of Salix petiolaris, the apparent sper- 
matogenous cells found in the archegonia of Mnium by Holferty 
('04), or the apparent megaspores found in the microsporangium 
of Marsilia by Shattuck ('10) are really capable of functioning 
as reproductive cells. 
We may now consider the case of Piper betel in the light of the 
observations just reviewed, to see whether these help in interpret- 
ing the facts recorded concerning this species, and also whether 
these facts support or controvert in any degree the views reached 
from the study of other forms. 
It is clear, from what has been said of Piper betel, that the sporo- 
phyte of this species is distinctly hermaphrodite, and also that the 
sexual character of each constituent of the hermaphrodite flower 
is already determined at the time of initiation of the stamens and 
carpels. It is likewise evident that the tissue of the young spike 
that is to bear perfect flowers must be potentially hermaphrodite 
in character. If this latter be true, then it seems probable that 
the tissue of those spikes, all of whose flowers are functionally 
unisexual, is likewise hermaphrodite at first. We must either 
admit this as proven by the fact that staminate flowers often bear 
some rudiments of megasporophylls, and the carpellate flowers 
nearly always bear rudimentary stamens, or else we must assume 
that, in the latter case, e.g., the male-determining substance is 
absent but the female-determining substance is capable of causing, 
or at least allowing, the development of the abortive stamen- 
like structures. Of these two possibilities, the evidence available 
seems to make the former view far more probable. In other words, 
it seems clear from the different degrees of suppression of micro- 
sporogenous tissue described in P. betel, that we must assume the 
male-determining substance to be present in all flowers, but more 
