This variety transmitted its characters so faithfully to all the succeeding self-fertilised generations, up to the last or ninth, that all the many plants which were raised presented a complete uniformity of character; thus offering a remarkable contrast with the seedlings raised from the purchased seeds.  Yet this variety retained to the last a latent tendency to produce yellow flowers; for when a plant of the eighth self-fertilised generation was crossed with pollen from a yellow-flowered plant of the Chelsea stock, every single seedling bore yellow flowers.  A similar variety, at least in the colour of its flowers, also appeared amongst the crossed plants of the third generation.  No attention was at first paid to it, and I know not how far it was at first used either for crossing or self-fertilisation.  In the fifth generation most of the self-fertilised plants, and in the sixth and all the succeeding generations every single plant consisted of this variety; and this no doubt was partly due to its great and increasing self-fertility.  On the other hand, it disappeared from amongst the crossed plants in the later generations; and this was probably due to the continued intercrossing of the several plants.  From the tallness of this variety, the self-fertilised plants exceeded the crossed plants in height in all the generations from the fifth to the seventh inclusive; and no doubt would have done so in the later generations, had they been grown in competition with one another.  In the fifth generation the crossed plants were in height to the self-fertilised, as 100 to 126; in the sixth, as 100 to 147; and in the seventh generation, as 100 to 137.  This excess of height may be attributed not only to this variety naturally growing taller than the other plants, but to its possessing a peculiar constitution, so that it did not suffer from continued self-fertilisation.

This variety presents a strikingly analogous case to that of the plant called the Hero, which appeared in the sixth self-fertilised generation of Ipomoea.  If the seeds produced by Hero had been as greatly in excess of those produced by the other plants, as was the case with Mimulus, and if all the seeds had been mingled together, the offspring of Hero would have increased to the entire exclusion of the ordinary plants in the later self-fertilised generations, and from naturally growing taller would have exceeded the crossed plants in height in each succeeding generation.

Some of the self-fertilised plants of the sixth generation were intercrossed, as were some in the eighth generation; and the seedlings from these crosses were grown in competition with self-fertilised plants of the two corresponding generations.  In the first trial the intercrossed plants were less fertile than the self-fertilised, and less tall in the ratio of 100 to 110.  In the second trial, the intercrossed plants were more fertile than the self-fertilised in the ratio of 100 to 73, and taller in the ratio of 100 to 92.  Notwithstanding that the self-fertilised plants in the second trial were the product of two additional generations of self-fertilisation, I cannot understand this discordance in the results of the two analogous experiments.