OBSEllVATIONS ON PROTANDRY AND PROTOGYNY. 319 



a strong tendency towards protandry ; in Stellaria graminea the versa- 

 tile anthers have dropped completely before the stigmas have become 

 papillose. In several species of Geranium the protandry is very 

 marked, in others not so decided. All the species of Leguminoscs 

 which I have examined present the same phenomenon, though not to 

 such a degree as to render self-fertilization impossible. In Bosacers we 

 find all grades, from the decided protandry of the Agrimony to the 

 marked protogyny of the Meadow-sweet. In the Bramble the enor- 

 mous crowd of stamens bending in a mass completely over the pistil, 

 and discharging their pollen while the stigmas are in a receptive con- 

 dition, would seem to imply self-fertilization ; and yet no flower is 

 more constantly and persistently infested by minute insects. In the 

 Ladies'-mantle {Alchemilla vulgaris) the stamens also bend over the 

 pistil while discharging their pollen ; but, on the other hand, the 

 stigmas do not rise up in a mature state tiU a later period. The Flax 

 family has furnished Mr. Darwin with some of his most striking in- 

 stances of dimorphism ; and yet in our little Linum catJtarticum , the 

 pistil and stamens are of the same length, and the pollen is discharged 

 quite on to the stigma, affording a privid facie instance of self-fertili- 

 zation. 



The Sundew [Drosera rotundifolia) is a little plant of very great 

 interest. Rarely opening its flowers, it yet almost invariably produces 

 its full complement of seed-bearing capsules. It appears to be strictly 

 synacmic, and I believe presents a very peculiar contrivance for en- 

 suring self-fertilization, which, however, I am not able fully to describe 

 on the present occasion. In Bipmcece, decided protandry appears to 

 be the rule. In the three species examined, the flower-heads present 

 two distinct phases ; as soon as the corolla opens, the large versatile 

 anthers appear at the end of their very slender filaments, the pistil 

 being quite hidden in the tube of the corolla ; after a time the anthers 

 drop off, and the filaments wither up, and then for tlie first time the 

 long styles and funnel-shaped stigmas appear, protruding considerably 

 beyond the mouth of the corolla. In Scabiosa succisa, and to a smaller 

 extent in other species of the Order, there is a singular combination, 

 which I do not remember to have seen noticed, of the centripetal and 

 centrifugal modes of inflorescence. The outer, larger, often rayed, 

 flowers of the capitulum open first, then those in the centre, leaving an 

 intermediate band which opens last of all. 



