204 ILLEGITIMATE OFFSPRING OF CHAP. V. 



suits of sowing seed naturally produced, there is reason 

 to believe that each form, when legitimately fertilised, 

 reproduces all three forms in about equal numbers. 

 Now, as we have seen (and the fact is a very singular 

 one) that the fifty-six plants produced from the 

 long-styled form, illegitimately fertilised with pollen 

 from the same form (Classes I. and II.), were all long- 

 styled. The short-styled form, when self-fertilised 

 (Class III.), produced eight short-styled and one long- 

 styled plant ; and the mid-styled form, similarly treated 

 (Class IV.), produced three mid-styled and one long- 

 styled offspring; so that these two forms, when ille- 

 gitimately fertilised with pollen from the same form, 

 evince a strong, but not exclusive tendency to repro- 

 duce the parent-form. When the short-styled form 

 was illegitimately fertilised by the long-styled form 

 '(Class V.), and again when the mid-styled was illegiti- 

 mately fertilised by the long-styled (Class VI.), in 

 each case the two parent-forms alone were reproduced. 

 As thirty-seven plants were raised from these two 

 unions, we may, with much confidence, believe that it 

 is the rule that plants thus derived usually consist of 

 both parent-forms, but not of the third form. When, 

 however, the mid-styled form was illegitimately fer- 

 tilised by the longest stamens of the short-styled 

 '(Class VII.), the same rule did not hold good; for the 

 seedlings consisted of all three forms. The illegiti- 

 mate union from which these latter seedlings were 

 raised is, as previously stated, singularly fertile, and 

 the seedlings themselves exhibited no signs of sterility 

 and grew to their full height. From the consideration 

 of these several facts, and from analogous ones to be 

 given under Oxalis, it seems probable that in a state 

 of nature the pistil of each form usually receives, 

 through the agency of insects, pollen from the stamens 



