Pot 6 : 28 2/8 : 18 6/8 : 12. Pot 6 : 22 : 7 : 16 1/8. Pot 6 : -- : 12 4/8 : --. Pot 7 : 12 4/8 : 15 : --. Pot 7 : 24 3/8 : 12 3/8 : --. Pot 7 : 20 4/8 : 11 2/8 : --. Pot 7 : 26 4/8 : 15 2/8 : --. Pot 8 : 17 2/8 : 13 3/8 : --. Pot 8 : 22 6/8 : 14 5/8 : --. Pot 8 : 27 : 14 3/8 : --. Pot 9 : 22 6/8 : 11 6/8 : --. Pot 9 : 6 : 17 : --. Pot 9 : 20 2/8 : 14 7/8 : --. Pot 10 : 18 1/8 : 9 2/8 : 10 3/8. Pot 10 : 16 5/8 : 8 2/8 : 8 1/8. Pot 10 : 17 4/8 : 10 : 11 2/8. Crowded plants. Total : 605.38 : 329.50 : 198.50. In this table the average height of the twenty-eight Chelsea-crossed plants is 21.62 inches; that of the twenty-seven intercrossed plants 12.2; and that of the nineteen self-fertilised 10.44. But with respect to the latter it will be the fairest plan to strike out two dwarfed ones (only 4 inches in height), so as not to exaggerate the inferiority of the self-fertilised plants; and this will raise the average height of the seventeen remaining self-fertilised plants to 11.2 inches. Therefore the Chelsea-crossed are to the intercrossed in height as 100 to 56; the Chelsea-crossed to the self-fertilised as 100 to 52; and the intercrossed to the self-fertilised as 100 to 92. We thus see how immensely superior in height the Chelsea-crossed are to the intercrossed and to the self-fertilised plants. They began to show their superiority when only one inch high. They were also, when fully grown, much more branched with larger leaves and somewhat larger flowers than the plants of the other two lots, so that if they had been weighed, the ratio would certainly have been much higher than that of 100 to 56 and 52. The intercrossed plants are here to the self-fertilised in height as 100 to 92; whereas in the analogous experiment given in Table 3/19 the intercrossed plants from the self-fertilised plants of the sixth generation were inferior in height to the self-fertilised plants in the ratio of 100 to 110. I doubt whether this discordance in the results of the two experiments can be explained by the self-fertilised plants in the present case having been raised from spontaneously self-fertilised seeds, whereas in the former case they were raised from artificially self-fertilised seeds; nor by the present plants having been self-fertilised during two additional generations, though this is a more probable explanation. With respect to fertility, the twenty-eight Chelsea-crossed plants produced 272 capsules; the twenty-seven intercrossed plants produced 24; and the seventeen self-fertilised plants 17 capsules. All the plants were left uncovered so as to be naturally fertilised, and empty capsules were rejected. Therefore 20 Chelsea-crossed plants would have produced 194.29 capsules. Therefore 20 Intercrossed plants would have produced 17.77 capsules. Therefore 20 Self-fertilised plants would have produced 20.00 capsules.