HETEROSTYLY 49 



in it the reason for the illegitimacy of some crossings : ' Since by further natural 

 selection the size of the pollen-grain has been adapted to the distance to be 

 traversed by the pollen-tube in legitimate crossing, and the stigmatic papillae have 

 been adapted to the size of pollen-grain they have to receive, it follows that sexual 

 organs standing at dissimilar heights would be unsuited for one another, and so 

 the illegitimate crossings of heterostylous flowers would be unfruitful.' 



C. Correns remarks (Ber. D. bot. Ges., Berlin, vii, 1889) that according 

 to this view there should really be three kinds of legitimate fertilization, i.e. $ -f 

 with 5 -f- , $ + with J -I- , and 5 -f with S H- ; but only one kind of 

 illegitimate fertilization, i. e. 5 -h with 5 -1- . Strasburger also (Jahrb. wiss. Bot., 

 Leipzig, xvii, 1886, p. 84) points out that the illegitimate crossing of long-styled 

 flowers among primroses (in which the small pollen-grain reaches the stigma of the 

 long style) should not, on this view, be more fertile than that of short-styled flowers. 



Naegeli (' Mechan. physiol. Theorie der Abslammungslehre,' p. 151) does not 

 consider that the diff"erent sizes of pollen-grains result from adaptation to the length 

 of the style, for the grains are nourished by the tissue of the style while developing 

 pollen-tubes. He sees in the differences in size and occasionally in colour of 

 the pollen-grains, merely an external sign of internal differentiation conditioning 

 illegitimacy; they perhaps depend, like the length of stigmatic papillae, upon the 

 height of insertion. 



Correns (op. cit.) proved by culture-experiments with the pollen of Primula 

 acaulis /acq., that the size of the pollen-grains has nothing to do either with the 

 distance to be traversed by the pollen-tube or with the cause of illegitimacy. The 

 following are the most important conclusions at which he arrived : 



1 . Both forms of pollen produce equally long tubes in the same time. 



2. The large pollen-grains develop thicker tubes than the small ones. 



3. The size of the pollen-grains is not an adaptation to the length of style to be 

 traversed in cases of legitimate fertilization, nor is it the cause of diminished fertiUty 

 in illegitimate crosses. 



4. There are no diff"erences discoverable in the capacity for absorbing nourish- 

 ment or in the chemotactic irritability, such as might explain the legitimacy or 

 illegitimacy of particular combinations. 



5. The length and form of the stigmatic papillae have likewise nothing to do 

 with the greater or less fertility of particular crosses. 



6. The lengths of the stigmatic papillae may be regarded as adapted to the 

 size of the grains, but only in the sense of facilitating their reception. 



Lastly, I would mention the interesting phenomenon of homo-heterostyly in 

 Menyanthes trifoliata. In Greenland, according to Warming, this otherwise always 

 dimorphic heterostylous plant is homostylous, the stigmas being at the same level as 

 the anthers. 



A list of the heterostylous plants hitherto recognized may be added: 



Trimorphous species: some Lythraceae, such as Lythrum Salicaria (Darwin). 



L. Graefferi (Darwin), L. virgatum, flexuosum, and maculatum (Kohne), Decodon 



verticillatus (Kohne), Nesaea Commers. and Lagerstroemia L. (Kuhn, Darwin), the 



linaceous Roucheria Planch. (? Kuhn), and twenty species of the genus Oxalis 



