single carpel, there being no trace of other carpels, and that this carpel resembles in structure and appearance one of the three that make up the tricarpellary pistil. In both cases the style is terminal and the ovary orthotropous at the beginning. As development takes place in the monocarpellary flower, one side of both ovary and ovule greatly outgrows the other so that the micropyle of the anatropous ovule faces the style, which appears to come from the base of the ovary. In the polycarpellary flower, the conjunction with other carpels pre- vents any one carpel bending to bring the style into a basal position. But the ovule becomes bent in the same direction as in the terminal flower, the funiculus being lifted and folded against the ovule by the elongation of the common axis of the three carpels. The result is a half-inverted or amphitropous ovule with the funiculus joining the common axis about one-third the distance from its top, and with the micropyle at the upper end and not far from the base of the style. It will be seen that development in both cases keeps the micropyle near the base of the style in the most favorable position for the ready entrance of the pollen tube.
The obturator, a mass of cells observed by Baillon (1) as growing from the placenta toward the micropyle of certain Euphorbiacese, is conspicuous here. We also observe the beak-like tip to the nucellus which was noticed by Lyon (2) in Euphorbia corollata and by Schweiger (3) and Weniger (4) in other species of Euphorbia.
The difference in the shape of the ovary leads to interesting differences between the terminal and lateral seeds. On the lateral seed is a conspicuous white caruncle formed by the tirgescence of the outer coat and occupying mainly the space between the funiculus and the micropyle. On the terminal seed, since the funiculus almost touches the micropyle, there is almost no caruncle. On the other hand its raphe is longer, extending the entire length of the seed.
Ten seeds of each type were selected at random from five different plants and were carefully measured. Those from terminal fruits averaged 1.247 mm. long by 1.047 mm. wide, those from lateral fruits averaged 1.447 mm. long by 1.115 mm. wide. The proportion of length to width of the former lot is 1.2 : 1, that of the latter is 1.3 : 1. It is also observable that the thickness of the terminal seed is about equal to its breadth, while that of the lateral seed is considerably less than its breadth.
In the material examined two interesting sports were discovered — one being a double flower of the terminal type occupying a terminal position, and the other a similar flower occupying a lateral position and lower down than other lateral flowers of the usual type.
