by efficient insects, I cannot account for this fact, more especially as the varieties are believed to intercross in North Italy.  Plants raised from a cross between two varieties, differing only in the colour of their flowers, grew much taller and were under unfavourable conditions more vigorous than the self-fertilised plants; they also transmitted, when self-fertilised, their superiority to their offspring.  The many varieties of the common Pea (Pisum sativum), though growing in close proximity, very seldom intercross; and this seems due to the rarity in this country of the visits of bees sufficiently powerful to effect cross-fertilisation.  A cross between the self-fertilised individuals of the same variety does no good whatever to the offspring; whilst a cross between distinct varieties, though closely allied, does great good, of which we have excellent evidence.  The flowers of the Broom (Sarothamnus) are almost sterile if they are not disturbed and if insects are excluded.  The pollen from a distinct plant is more effective than that from the same flower in producing seeds.  The crossed seedlings have an enormous advantage over the self-fertilised when grown together in close competition.  Lastly, only four plants of the Ononis minutissima were raised; but as these were observed during their whole growth, the advantage of the crossed over the self-fertilised plants may, I think, be fully trusted.

[15.  Onagraceae.—­Clarkia elegans.

Owing to the season being very unfavourable (1867), few of the flowers which I fertilised formed capsules; twelve crossed flowers produced only four, and eighteen self-fertilised flowers yielded only one capsule.  The seeds after germinating on sand were planted in three pots, but all the self-fertilised plants died in one of them.  When the two lots were between 4 and 5 inches in height, the crossed began to show a slight superiority over the self-fertilised.  When in full flower they were measured, with the following result:—­

Table 5/62.  Clarkia elegans.

Heights of plants measured in inches.

Column 1:  Number (Name) of Pot.

Column 2:  Crossed Plants.

Column 3:  Self-fertilised Plants.

Pot 1 :  40 4/8 :  33. 
Pot 1 :  35 :  24. 
Pot 1 :  25 :  23.

Pot 2 :  33 4/8 :  30 4/8.

Total :  134.0 :  110.5.

The average height of the four crossed plants is 33.5, and that of the four self-fertilised plants 27.62 inches, or as 100 to 82.  The crossed plants altogether produced 105 and the self-fertilised plants 63 capsules; or as 100 to 60.  In both pots a self-fertilised plant flowered before any one of the crossed plants.

16.  Loasaceae.—­Bartonia aurea.

Some flowers were crossed and self-fertilised in the usual manner during two seasons; but as I reared on the first occasion only two pairs, the results are given together.  On both occasions the crossed capsules contained slightly more seeds than the self-fertilised.  During the first year, when the plants were about 7 inches in height, the self-fertilised were the tallest, and in the second year the crossed were the tallest.  When the two lots were in full flower they were measured, as in Table 5/63.