THE NUMBER OF STAMENS AND CARPELS. 245 
of both organs, than those formed later in the season; and, 
again, we see that the flowers formed earliest, or holding any 
position of advantage on an inflorescence, 7. e. terminal on a 
cyme or at the base of a raceme, eveu if not maturing earliest, 
e.g. Thalictrum, carry more stamens or carpels, or more of both, 
than the other flowers of the same inflorescence. Stellaria media 
and Ranunculus Ficaria, where the early and late flowers are 
separated from each other by a considerable interval of time, 
show the reduction in later flowers, not necessarily on the same 
axes as the earlier ones. 
The other plants examined show how in the same inflorescence 
we can appreciate a reduction—now in the stamens, now in the 
carpels—in the later-formed flowers. What other influences act 
is a question too involved to answer other than conjecturally. 
I have suggested temperature in the case of Stellaria and Ranun- 
culus bulbosus, but am not able to hazard any other cause with 
any degree of safety. What experiments follow will, I hope, 
` throw further light upon this tangled question. 
Apart from these considerations, I have been able to show that 
the order of abortion in the stamens of Stellaria media is fixed 
and definite; and that in several plants there exists a correlation 
between the numbers of petals and stamens and of stamens and 
carpels—a higher average number of one organ being present, 
there will be a corresponding increase in the other organs. 
In conclusion, I desire to express my thanks to Mr. Francis 
Darwin for much kindly advice, and to Mr. H. F. Newall for the 
data of temperature and rainfall with which he supplied me. To 
Mr. R. I. Lynch I am indebted for many services; and, above all, 
my thanks are due to Mr. J. C. Willis, to whom I owe the 
original suggestion in regard to these investigations, and who 
has moreover, rendered me considerable assistance. 
LINN. JOURN.— BOTANY, VOL. XXXI. 
