212 ILLEGITIMATE OFFSPRING OP Chap. V. 



two distinct species, for the seeds in this latter case 

 are often ijicapable of germination. 



The foUowtDg observations relate to the nature of the forms 

 which appear among the legitimate seedlings of Oxdlis Valdiviana. 

 Hildebrand raised, as described in the paper just referred to, 

 211 seedlings from all six legitimate unions, and the three forms 

 appeared among the offspring from each union. For instance, 

 long-styled plants were legitimately fertilised with pollen from 

 the longest stamens of the mid-styled form, and the seedlings 

 consisted of 15 long-styled, 18 mid-styled, and 6 short-styled. 

 We here see that a few short-styled plants were produced, though 

 neither parent was short-styled; and so it was with the other 

 legitimate unions. Out of the above 211 seedlings, 173 belonged 

 to the same two forms as their parents, and only 38 belonged 

 to the third form distinct from either parent. In the case of 

 0. BegnelU, the result, as observed by Hildebrand, was nearly 

 the same, but more striking : aU the offspring from four of the 

 legitimate unions consisted of the two parent-forms, whilst 

 amongst the seedlings from the other two legitimate unions the 

 third form appeared. Thus, of the 43 seedhngs from the six 

 legitimate unions, 35 belonged to the same two forms as their 

 parents, and only 8 to the third form. Fritz Miiller also raised 

 in Brazil seedlings from long-styled plants of 0. SegnelU legiti- 

 mately fertilised with poUen from the longest stamens of the 

 mid-styled form, and all these belonged to the two parent- 

 forms.* Lastly, seedlings were raised by me from long-styled 

 plants of 0. spedosa legitimately fertilised by the short-styled 

 form, and from the latter reciprocally fertilised by the long- 

 styled; and these consisted of 33 long-styled and 26 short- 

 styled plants, with not one mid-styled form. There can, there- 

 fore, be no doubt that the legitimate offspring from any two 

 forms of OxaUs tend to belong to the same two forms as their 

 parents ; but that a few seedlings belonging to the third form 

 occasionally make their appearance ; and this latter fact, as 

 Hildebrand remarks, may be attributed to atavism, as some of 

 their progenitors wiU almost certainly have belonged to the 

 third form. 



When, however, any one form of Oxalis is fertilised illegiti- 



* ' Jenaisohe Zeitechrift,' &c. Band vi. 1871, p. 75. 



