ENE saps 
ae Se eee 
ae en 
pantie sen iedreos: 
1898 ] COMPARATIVE MORPHOLOGY OF PISTILS 309 
less like Ranunculus than is Sagittaria, for the ovule is hardly 
axillary with respect to the carpel, but arises from the recep- 
tacle, in this respect much resembling the origin of the sporan- 
gium of Selaginella. The condition found in Sagittaria is one 
about midway between that in Alisma and that in Ranunculus. 
In the latter the axillary placenta is in reality only an outgrowth 
from the receptacle, and this prepares us to find (as in the Alis- 
macez) the ovule developed directly from the receptacle. In 
other cases, as in Potentilla, this axillary placenta loses its indi- 
viduality by fusion with one of the lamin of the carpel. 
The presence of the accessory ovules in Anemone, Pulsatilla, 
Clematis, and other genera is difficult to explain. If it were not 
for the peculiar origin of the one ovule which reaches maturity, 
it might naturally be supposed that Anemone is descended from 
plants with multiovulate pistils. However, if this were the case 
it would be necessary to consider also that Ranunculus and 
Myosurus had a similar origin, which seems highly improbable 
in consideration of their close resemblance to the Alismacee, 
which show in other characters no signs of having had Anemone- 
like ancestors. Furthermore, there are no existing multiovulate 
Ranunculacee that seem to be as simple in other respects as 
Ranunculus, for their pistils are fewer in number and close much 
earlier, an evident unsimilarity to the theoretical pteridophytic 
ancestors of the angiosperms. Perhaps the best solution of the 
problem is the supposition that some plant of the Ranunculus or 
Myosurus type after the development of its first ovule varied so 
as to develop in the space above the ovule one or more acces- 
sory ovules which were unable to reach maturity. These acces- 
sory ovules being in the unoccupied upper part of the pistil out 
of the way of the large ovule, and yet protected by the carpel 
wall, would have no part in the struggle of the plant for exist- 
ence, and so might persist. This would be the more likely to be 
true if this modification happened to occur in a plant which, 
owing to other modifications, was enabled to maintain itself 
against all enemies, and to be well distributed. This seems ’ 
have been the case here, for these accessory ovules are found in 
