94 WILLIAM TRELEASE ON THE . 



replacing flowers, and Plukenet^ states that the flowers of our O. violacea are sometimes 

 succeeded by bulblets, — an oeciiiTence that cannot be common in the wikl plant. 



The flowers are typically open funnel-shaped, expanding in sunshine, the degree of 

 warmth playing a very imjiortant part in this jii'ocess, at least in some species^ The 

 stamens are in two sets of unequal length, as in Geranium, etc., Avhile the styles appear 

 to be typically about equal to the longer set. But in the majority of species the flowers 

 of different individuals are quite dissimilar in this resjiect, so that these species are what 

 Hildebrand'' and Darwin* have called heterostyled, or Gray^, heterogone. In Oxalis the 

 heterogony is typically trimorphism, three forms of flowers existing in about equal num- 

 bers on difterent plants of a given species, the pistils of one being shorter than the short- 

 est stamens (short-styled), those of the second being intermediate between the two sets 

 (mid-styled), while those of the third are longer than either set of stamens (long-styled) . 

 Connected with these differences in the length of the pistil are differences in the stamens, 

 so that the two sets of either form of flower correspond in length res2:)ectively with the 

 jjistils of the two other forms. Hildelirand, who examined the sjDecimens of Oxalis in 

 several large herbaria, has recorded his observations on this point in detail*'. But while 

 the appearance in a majority of cases was that of trimorphism, only one or two forms of 

 flowers were found in many of the species examined. In most instances, however, these 

 correspond to one of the three forms that I have characterized. The same may be said 

 of the man}' figures of Jacquin" and others who have illustrated this genus. It Avas also 

 shown by Hildebrand^ that in some trimorphic species submitted to experiment, the gi'eat- 

 est fruitfulness resulted only from the fertilization of a pistil by the pollen from stamens 

 of its own length, — a rule that has been found to a})ply generally to heterogone plants'*; 

 and Hildebrand has recently again remarked on the frequent self-sterility of a single 

 form^". The pollen and stigmas of such trimorphic species as have been examined in this 

 respect, also show the differentiation usual in trimorphic plants, the pollen grains of the 

 long stamens being largest, those of the short stamens, smallest, and those of the mid- 

 length, intermediate in size; while the stigmatic papillae of the long-styled flowers are 

 longest, those of the short-styled, shortest, and those of the mid-styled, of intermediate 

 length. 



In the examination of herbarium or greenhouse material of trimorphic species, the nimi- 

 ber of specimens being limited, it may readily happen that only one or tAvo of the forms 

 are found, when a search in the field among hundreds of individuals would reveal all in 

 approximately equal numbei-s. But even under these conditions, one or more of the 

 forms may not be found in localities where an introduced species has multiplied non-sex- 

 ually", or where but two forms have been introduced; for these, propagated sexually, but 

 without the intervention of the other forms, may possibly reproduce only themselves^^. 

 Hence the presumable trimorphism of such species as were known to have one or two of 

 the forms of flowers usual in such cases has not been generallj^ disputed. 



Some years since, having access for the first time to numerous wild plants of a sup- 



' Almagestum Bot. i, PI. 102, dg. 4. « Bot. Zeituiig, 1871, 415, 431. 



^ Cf. Pfeffer: Physiol. Uiiters. 1873. ' Darwin: Diflferent Forms of Flowers, etc., Chapter 5. 



■' Monatsber. Berlin. Akad. ISGfi, 3fi9. '" Ilildebraucl: Bot. Zeit. 1887, iios. 1-3. 



■* Different Forms of Flowers, various places. " Darwin : Different Forms of Flowers, 182. 



* Amer. Journ. Soi. and Arts, 3 Ser., xiii, 82. '- llildeljrand: Monatsl)er. Berl. Akad., 18fifi, 373; Bot. 



^ Monatsber. Berlin. Al<ad., 1806,351. Zeit., 1887, 5 ;Darwin: Different Forms of Flowers, 268. 



' Oxalis Monograpliia, 1794. 



