the seed to its base gives way, and it is ejected several feet, this being no doubt much facilitated by its form and smoothness. I have known even a gathered specimen throw a seed nearly ten feet. Fig. 5 represents a capsule after the seeds have been ejected.
Now, we naturally ask ourselves what is the reason for this difference between the species of violets; why do Viola odorata and Viola Hirta conceal their capsules among the moss and leaves on the ground, while Viola canina and others raise theirs boldly above their heads, and throw the seeds to seek their fortune in the world? If this arrangement be best for Viola canina, why has not Viola odorata also adopted it? The reason is, I believe, to be found in the different mode of growth of these two species. Viola canina is a plant with an elongated stalk, and it is easy, therefore, for the capsule to raise itself above the grass and other low herbage among which violets grow.
Fig. 6.—The Herb-robert (Geranium Robertianum,) a, bud; b, flower; c, flower after the petals have fallen; d, flower with seeds nearly ripe; e, flower with ripe seeds; f, flower after throwing seeds.
Viola odorata and Viola hirta, on the contrary, have, in ordinary parlance, no stalk, and the leaves are radical, i. e., rising from the root. This is at least the case in appearance, for, botanically speaking, they rise at the end of a short stalk. Now, under these circumstances, if
