the degree of self-fertilisation was not quite so close as in the last generation, in which pollen from the same flower, kept in paper, was used.  These two lots of seeds were thinly sown on opposite sides of nine pots; and the young seedlings were thinned, an equal number of nearly as possible the same age being left on the two sides.  In the spring of the following year (1870), when the seedlings had grown to a considerable size, they were measured to the tips of their leaves; and the twenty-three crossed plants averaged 14.04 inches in height, whilst the twenty-three self-fertilised seedlings were 13.54 inches; or as 100 to 96.

In the summer of the same year several of these plants flowered, the crossed and self-fertilised plants flowering almost simultaneously, and all the flower-stems were measured.  Those produced by eleven of the crossed plants averaged 30.71 inches, and those by nine of the self-fertilised plants 29.43 inches in height; or as 100 to 96.

The plants in these nine pots, after they had flowered, were repotted without being disturbed in much larger pots; and in the following year, 1871, all flowered freely; but they had grown into such an entangled mass, that the separate plants on each side could no longer be distinguished.  Accordingly three or four of the tallest flower-stems on each side of each pot were measured; and the measurements in Table 5/71 are, I think, more trustworthy than the previous ones, from being more numerous, and from the plants being well established and growing vigorously.

The average height of the thirty-four tallest flower-stems on the twenty-three crossed plants is 29.82 inches, and that of the same number of flower-stems on the same number of self-fertilised plants is 27.10 inches, or as 100 to 91.  So that the crossed plants now showed a decided advantage over their self-fertilised opponents.

22.  Polemoniaceae.—­Nemophila insignis.

Twelve flowers were crossed with pollen from a distinct plant, but produced only six capsules, containing on an average 18.3 seeds.  Eighteen flowers were fertilised with their own pollen and produced ten capsules, containing on an average 12.7 seeds, so that the seeds per capsule were as 100 to 69. (5/25.  Several species of Polemoniaceae are known to be proterandrous, but I did not attend to this point in Nemophila.  Verlot says ‘Des Variétés’ 1865 page 66, that varieties growing near one another spontaneously intercross.) The crossed seeds weighed a little less than an equal number of self-fertilised seeds, in the proportion of 100 to 105; but this was clearly due to some of the self-fertilised capsules containing very few seeds, and these were much bulkier than the others, from having been better nourished.  A subsequent comparison of the number of seeds in a few capsules did not show so great a superiority on the side of the crossed capsules as in the present case.