Primula veris—­(equal-styled, red-flowered variety)—­offspring from plants self-fertilised for two generations and then crossed by a different variety, compared with plants of the third self-fertilised generation, in fertility: 
  .. :  .. :  .. :  .. :  11.

In these three tables the measurements of fifty-seven species, belonging to fifty-two genera and to thirty great natural families, are given.  The species are natives of various parts of the world.  The number of crossed plants, including those derived from a cross between plants of the same stock and of two different stocks, amounts to 1,101; and the number of self-fertilised plants (including a few in Table 7/C derived from a cross between plants of the same old stock) is 1,076.  Their growth was observed from the germination of the seeds to maturity; and most of them were measured twice and some thrice.  The various precautions taken to prevent either lot being unduly favoured, have been described in the introductory chapter.  Bearing all these circumstances in mind, it may be admitted that we have a fair basis for judging of the comparative effects of cross-fertilisation and of self-fertilisation on the growth of the offspring.

It will be the most convenient plan first to consider the results given in Table 7/C, as an opportunity will thus be afforded of incidentally discussing some important points.  If the reader will look down the right hand column of this table, he will see at a glance what an extraordinary advantage in height, weight, and fertility the plants derived from a cross with a fresh stock or with another sub-variety have over the self-fertilised plants, as well as over the intercrossed plants of the same old stock.  There are only two exceptions to this rule, and these are hardly real ones.  In the case of Eschscholtzia, the advantage is confined to fertility.  In that of Petunia, though the plants derived from a cross with a fresh stock had an immense superiority in height, weight, and fertility over the self-fertilised plants, they were conquered by the intercrossed plants of the same old stock in height and weight, but not in fertility.  It has, however, been shown that the superiority of these intercrossed plants in height and weight was in all probability not real; for if the two sets had been allowed to grow for another month, it is almost certain that those from a cross with the fresh stock would have been victorious in every way over the intercrossed plants.

Before we consider in detail the several cases given in Table 7/C, some preliminary remarks must be made.  There is the clearest evidence, as we shall presently see, that the advantage of a cross depends wholly on the plants differing somewhat in constitution; and that the disadvantages of self-fertilisation depend on the two parents, which are combined in the same hermaphrodite flower, having a closely similar constitution.  A certain amount of differentiation in the sexual