48 INTRODUCTION 



1. The legitimate unions usually brought about by the normal flower-visitors 

 yield, almost without exception, a larger quantity of seed than illegitimate unions. 



2. The difference as to fertility between legitimate and illegitimate unions varies 

 in different species. Within the same genus there are some species in which 

 illegitimate unions are almost entirely sterile (Oxalis Valdiviana, O. Regnelli), as well 

 as others in which they are comparatively fertile (O. speciosa). 



3. The various legitimate unions vary among themselves in respect of fertility, 

 either to a small extent only (Oxalis Valdiviana, O. Regnelli) or in a higher degree 

 (Lythrum Salicaria). Among dimorphous plants the union of short-styled forms with 

 pollen from long-styled ones is as a rule the more fertile, but the opposite is true for 

 Hottonia and Primula acaulis. Among trimorphous species the relations are yet more 

 varied : in the case of Lythrum Salicaria the legitimate unions of the medium-styled 

 form are the most prolific. Then comes, as regards seed production, the long-styled 

 form, and last of all the short-styled form. In other trimorphous species, e. g. Oxalis 

 speciosa, the sequence is not the same, and the differences between the several cases 

 are not so great as in Lythrum. 



4. Among the illegitimate unions of dimorphous species, those of the long-styled 

 form are as a rule more fertile than those of the short-styled one (except, e. g., 

 Pulmonaria, where Hildebrand found that both illegitimate unions were quite sterile). 

 The illegitimate unions of a trimorphous species present relatively greater differences 

 than those of dimorphous species. In Lythrum Salicaria, e.g., the illegitimate unions 

 of the medium-styled form are conspicuously fertile, while, on the contrary, those 

 of the other two forms are very infertile. According to Darwin, the infertility is 

 here directly proportionate to the difference in length between the pistil and the 

 filaments of the stamens which furnish the pollen for fertilization. In Oxalis speciosa 

 some illegitimate unions of all three forms are fertile to a certain degree, but those of 

 the medium-styled flowers are not more fertile than the rest. 



Correns has pointed out that F. Delpino suggested an explanation ('La Distribu- 

 zione dei sessi nelle piante,' 1867, p. 17) of the difference between the two forms of 

 dimorphous heterostylic flowers as to the size of their pollen-grains. His view was 

 that the greater volume of the pollen-grains of the short-styled flowers ( 5 -f- ) ' 

 has reference to the longer distance that the pollen-tubes have to travel in cases 

 of legitimate fertilization, while the pollen-grains of the long-styled flowers ( ? -f- ) ^ 

 are smaller because their tubes have a shorter distance to travel in legitimate 

 fertilization. 



Charles Darwin ('Forms of Flowers,' pp. 250 et seq.) has likewise given no 

 sufficient answer to this question. He was not ignorant of the fact that difference 

 in size of pollen-grains does not appear in all dimorphous heterostylic flowers (e. g. 

 species of Linum), and that, conversely, when there is marked difference of size 

 in the pollen-grains there may be very slight differences as to length of styles 

 (e.g. Suteria). Darwin supposed that the pollen-grain is sometimes nourished 

 by the tissue of the style. 



Hermann Miiller (' Wechselbeziehungen,' p. 86) regards the difference in size 

 of the pollen-grains as an adaptation to the length of the style, and even recognizes 



' The + sign is used to suggest the larger pollen-grains and longer style, respectively. Tr. 



