If we now consider the crossed and self-fertilised plants of the three generations, we find an extraordinary diversity in their relative heights.  In the first generation, the crossed plants were inferior to the self-fertilised as 100 to 178; and the flowers on the original parent-plants which were crossed with pollen from a distinct plant yielded much fewer seeds than the self-fertilised flowers, in the proportion of 100 to 150.  But it is a strange fact that the self-fertilised plants, which were subjected to very severe competition with the crossed, had on two occasions no advantage over them.  The inferiority of the crossed plants of this first generation cannot be attributed to the immaturity of the seeds, for I carefully examined them; nor to the seeds being diseased or in any way injured in some one capsule, for the contents of the ten crossed capsules were mingled together and a few taken by chance for sowing.  In the second generation the crossed and self-fertilised plants were nearly equal in height.  In the third generation, crossed and self-fertilised seeds were obtained from two plants of the previous generation, and the seedlings raised from them differed remarkably in constitution; the crossed in the one case exceeded the self-fertilised in height in the ratio of 100 to 83, and in the other case were almost equal.  This difference between the two lots, raised at the same time from two plants growing in the same pot, and treated in every respect alike, as well as the extraordinary superiority of the self-fertilised over the crossed plants in the first generation, considered together, make me believe that some individuals of the present species differ to a certain extent from others in their sexual affinities (to use the term employed by Gartner), like closely allied species of the same genus.  Consequently if two plants which thus differ are crossed, the seedlings suffer and are beaten by those from the self-fertilised flowers, in which the sexual elements are of the same nature.  It is known that with our domestic animals certain individuals are sexually incompatible, and will not produce offspring, although fertile with other individuals. (6/3.  I have given evidence on this head in my ‘Variation of Animals and Plants under Domestication’ chapter 18 2nd edition volume 2 page 146.) But Kolreuter has recorded a case which bears more closely on our present one, as it shows that in the genus Nicotiana the varieties differ in their sexual affinities. (6/4.  ‘Das Geschlecht der Pflanzen, Zweite Fortsetzung’ 1764 pages 55-60.) He experimented on five varieties of the common tobacco, and proved that they were varieties by showing that they were perfectly fertile when reciprocally crossed; but one of these varieties, if used either as the father or the mother, was more fertile than any of the others when crossed with a widely distinct species, N. glutinosa.  As the different varieties thus differ in their sexual affinities, there is nothing surprising in the individuals of the same variety differing in a like manner to a slight degree.