White flowers, in the opinion of the writer, are due to retrogression, and are derived from yellow, red, or blue, and in some instances from the primitive green, as in the involucre of _Cornus_. As a whole they present no advance in specialization over yellow flowers, and are often smaller and less conspicuous. When the petals of blossoms containing yellow, red, or blue pigments are placed in concentrated alcohol they turn to white. To produce these pigments is evidently more or less a tax upon the energies of the plant, which, whenever possible, is avoided. They are not present in the embryonic buds, and may not develop until they are well advanced in size. In _Gentiana crinita_ the yellowish-white bud is nearly an inch long before the purple coloring appears, and the corolla always remains white at base. A stimulus to the growth of the plant makes itself apparent in the increased brilliancy of the flowers, as when they are exposed to clear sunlight or are treated with nitrate of soda, and may also be observed in the flushing of tulips, by which they lose their variegated colors when treated with strong manure. On the other hand, a check in nutrition and growth will cause a diminution of the perianth in size, accompanied by retrogression in color. When double red poppies are transplanted the whole plant is dwarfed, while the flowers are much smaller and pure white. This view of the origin of white flowers explains why they are the commonest in Nature, accounts for their being most numerous in families in which yellow flowers are likewise numerous, and why they are most true to name under cultivation. Many white flowers also exhibit other evidences of degeneration in their structure. Numerous species of _Cruciferæ_ and _Caryophyllaceæ_ have small white flowers, which regularly fertilize themselves; and in _Lepidium_, _Stellaria_, and _Sagina_ the petals are sometimes present and in other instances are wanting.