Chap. V. HETEROSTYLED TKIMOEPHIO PLANTS. 207 



The lessened fertility of most of these illegitimate 

 plants is in many respects a highly remarkabl© phe- 

 nomenon. Thirty-three plants in the seven classes 

 were subjected to various trials, and the seeds care- 

 fully counted. Some of them were artificially ferti- 

 lised, but the far greater number were freely fertilised 

 (and this is the better and natural plan) through the 

 agency of insects, by other illegitimate plants. In the 

 right-hand, or percentage column, in the preceding 

 table, a wide difference in fertility between the plants 

 in the first four and the last three classes may be per- 

 ceived. In the first four classes the plants are de- 

 scended from the three forms illegitimately fertilised 

 with pollen taken from the same form, but only 

 rarely from the same plant. It is necessary to observe 

 this latter circumstance; for, as I have elsewhere 

 shown,* most plants, when fertilised with their own 

 pollen, or that from the same plant, are in some 

 degree sterile, and the seedlings raised from such 

 unions are likewise in some degree sterile, dwarfed, 

 and feeble. None of the nineteen illegitimate plants 

 in the first four classes were completely fertile ; one, 

 however, was nearly so, yielding 96 per cent, of the 

 proper number of seeds. From this high degree of 

 fertility we have many descending gradations, till we 

 reach an absolute zero, when the plants, though bear- 

 ing many flowers, did not produce, during successive 

 years, a single seed or even seed-capsule. Some of the 

 most sterile plants did not even yield a single seed 

 when legitimately fertilised with pollen from legiti- 

 mate plants. There is good reason to believe that the 

 first seven plants in Class I. and II. were the offspring- 



* 'The Eifpcts of Cross and Self-fertilisation in the Vegetable 

 Kiujidom,' 1876. 



