Buttercups and Columbines 61 
anthers successively mature, they turn as far away 
from the stigmas as possible, so that they have every 
chance of being dusted by pollen shaken from the 
bodies of visitors who have been to other flowers 
previously. By the wandering of the smaller insects 
from anthers to stigmas, there is no doubt that 
fertilisation is frequently effected. The carpels 
develop into achenes not much unlike those of the 
Potentillas; but it has other means of reproduction, 
for frequently little tubers form at the base of the 
leaf-stalks, and as the plant fades these strew the 
ground in places, and each produces a new plant next 
year. As these tubers are about the size of grains of 
wheat, their presence on the ground has sometimes led 
people to imagine that it has rained wheat! The 
swellings on the roots are also detachable, and each 
capable of producing a new plant. 
It is interesting to consider the probable cause of 
the tuberous roots in Celandine. There can be little 
doubt that it is one of the oldest forms, if not the 
oldest form of Ranunculus we have—the shape of the 
lower leaves testifies to that. Perhaps Celandine and 
Serpent’s-tongue (Renunculus ophioglossifolius)— 
now all but extinct in this country—exhibit the 
branching off from a common ancestral form, the 
former towards Marsh Marigold and Globe-flower, 
the latter to the Buttercups through the Spearworts. 
Celandine was then probably an annual, as Serpent’s- 
tongue is still, and flowered in summer, as most of the 
family do. We can imagine Celandine growing as 
now under the shelter of bushes, but not being so 
pressed by competitors for space and light. Some 
changes brought about the crowding of such places by 
