A capsule was obtained by crossing a flower of L. ramosa with pollen from another plant, and two other capsules from artificially self-fertilised flowers. The contained seeds were sown on the opposite sides of four pots. Some of the crossed seedlings which came up before the others had to be pulled up and thrown away. Whilst the plants were very small there was not much difference in height between the two lots; but in Pot 3 the self-fertilised were for a time the tallest. When in full flower the tallest plant on each side of each pot was measured, and the result is shown in Table 5/68. In all four pots a crossed plant flowered before any one of its opponents. TABLE 5/68. Lobelia ramosa (First Generation). Heights of plants measured in inches. Column 1: Number (Name) of Pot. Column 2: Tallest Crossed Plant in each Pot. Column 3: Tallest Self-fertilised Plant in each Pot. Pot 1 : 22 4/8 : 17 4/8. Pot 2 : 27 4/8 : 24. Pot 3 : 16 4/8 : 15. Pot 4 : 22 4/8 : 17. Total : 89.0 : 73.5. The four tallest crossed plants averaged 22.25, and the four tallest self-fertilised 18.37 inches in height; or as 100 to 82. I was surprised to find that the anthers of a good many of these self-fertilised plants did not cohere and did not contain any pollen; and the anthers even of a very few of the crossed plants were in the same condition. Some flowers on the crossed plants were again crossed, four capsules being thus obtained; and some flowers on the self-fertilised plants were again self-fertilised, seven capsules being thus obtained. The seeds from both lots were weighed, and it was calculated that an equal number of capsules would have yielded seed in the proportion by weight of 100 for the crossed to 60 for the self-fertilised capsules. So that the flowers on the crossed plants again crossed were much more fertile than those on the self-fertilised plants again self-fertilised. PLANTS OF THE SECOND GENERATION.