Chap. V. HETEROSTYLED TRIMORPHIC PLANTS. 209 



the seven classes, the plants, though descended from 

 the same parents, sown at the same time and in the 

 same soil, differed much in their average degree of 

 fertilitv. 



Turning now to the fifth, sixth, and seventh classes, 

 and looking to the right-hand column of the table, we 

 find nearly as many plants with a percentage of seeds 

 above the normal standard -.as beneath it. As with 

 most plants the number of seeds produced varies much, 

 it might be thought that the present case was one 

 merely of variability. But this view must be rejected 

 as far as the less fertile plants in these three classes 

 are concerned: first, because none of the plants in 

 Class V. attained their proper height, which shows 

 that they were in some manner affected; and secondly, 

 because many of the plants in Classes V. and VI. pro- 

 duced anthers which were either confcabescent or in- 

 cluded small and shrivelled pollen-grains. And as in 

 these cases the male organs were manifestly deterio- 

 rated, it is by far the most probable conclusion that 

 the female organs were in some cases likewise affected, 



and that this was the cause of the reduced number of 

 seeds. 



With respect to the six plants in these three classes 

 which yielded a very high percentage of seeds, the 

 thought naturally arises that the normal standard of 

 fertility for the long-styled and short-styled forms 

 (with which alone we are here concerned) may have 

 been fixed too low, and that the six illegitimate plants 

 are merely fully fertile. The standard for the long- 

 styled form was deduced by counting the seeds in 

 twenty-three capsules, and for the short-styled form 

 from twenty-five capsules. I do not pretend that this 

 is a sufficient number of capsules for absolute accu- 

 racy; but my experience has led me to believe that a 



