498 



NA TURE 



[Oct. 17, 1872 



ON THE FERTILISATION OF A FEW COM- 

 MON PAPILIONACEOUS FLOWERS 

 II. 

 J/ICIA SATIVA. — In the general structure and cha- 

 ^ racter of petals, stamens, and pistil, this flower 

 agrees with Lathyrus ; but there is a remarkable diffe- 

 rence in the shape of the keel, and correlatively in the 

 hairs or brush on the style. 



In Lathyrus, as we have seen, the upper part of the 

 elastic style is curved, so that the curvature corresponds 

 with the curvature of the keel ; the back or outside of the 

 style presses against the inside of the keel, and is not 

 furnished with hairs, there being no space for pollen on 

 that side, whilst the inside of the style is covered with 

 hairs set upwards, so as to sweep out the pollen which 

 accumulates on this side. 



In Vicia sath'a the keel forms a less regular curve, 

 whilst the style, instead of following the curvature of the 

 keel, is set on at right angles to the ovary, and is straight 

 and perpendicular throughout its whole length. There 

 is, therefore, a large nook or corner outside the style, and 

 between it and the keel, into which the pollen gets. Cor- 

 relatively the style is not furnished with abundant hairs 

 on the inside, as in Lathyrus ; but there is a little tuft of 

 stiff hairs on the outside, a little below the stigma, set on 

 upwards so as exactly to brush out the pollen from the 

 nook of the keel, when the keel is pressed down by an 

 insect (see Fig. 9). 



Fk,. 9.-/-, 



ati-ja (ktel and pistil). 



Vicia sepium is similar in construction to V. sativa. 

 I have not observed whether there is nectar within the 

 st.aminal tube of V. saliva or V. sepium, but feel assured 

 that it is to be found there. 



I'iciaFaba. In the several positions of its buds and 

 pods, Faba (broad bean) differs from Pisum and Lathyrus, 

 and agrees with Phaseolus. The buds are upright ; in 

 the flower they are horizontal, and in the pod they are 

 again upright ; but the blossom, when open, is, as in all 

 the other cases, horizontal, so as to afford a good lighting 

 place for bjes which seek the nectar in the interior of the 

 staminal tube. This tube, the separate stamen, the shape 

 of the keel of the style with its brush, are similar to those 

 of V. sativa. 



Rohinia pscud-Acacia. — This plant bears its flowers 

 in a pendent raceme ; consequently, the position of the 

 flower is reversed. The fifth sepal should be uppermost, 

 and the vexiUum lowest, with its back to the peduncle ; 

 and this is the position of the unopened buds. But as 

 they approach maturity, ix. as the blossom opens, the 

 pedicel of each flower takes a half twist, so as to bring the 

 flower into what we may call the normal position of Papi- 

 lionacea;, but with the vexiUum uppermost and upright, 

 or nearly so, and the wings and keel horizontal, the open 

 side of the keel being uppermost. The keel is obtuse, 

 and is free from the wings. 



The filament of the tenth stamen is joined to the others 

 in the middle, with apertures between it and the others at 

 the base, and there is a cavity at the base of the staminal 

 tube containing nectar. The stigma has a very small 

 brush round it, and there are a few hairs on the inside of 

 the style which seem to sweep out the pollen. The 

 flowers are much frequented by bees. 



Wistaria sinensis. — The pendent raceme of this plant 



displays, as regards the position of the buds and flowers, 

 the same features as thatof AV/'/«/,;/jvm;'-^icrt«>;, whichit 

 also resembles in the free boat-shaped keel, the semi- 

 separated tenth stamen, and the nectar-holding cavity of 

 the staminal tube. It differs in having no hairs on the 

 style, a difference possibly connected with the character 

 of the pollen. But as the flower does not usually pro- 

 duce seed in this country, it seems unsafe to speculate on 

 such a point. 



Onobiycliis sativa, or Sainfoin. — In the long raceme of 

 thisplant,the pcdicclsof the flower are nearly perpendicular 

 in the bud, horizontal, as usual, in the flower, and again, 

 after flowering, resume a position approaching the perpen- 

 dicular. The wings are very small, and are rxOt attached 

 to the keel, and seem to play no part in fertilisation. On 

 the other hand, the keel is large, boat-shaped, prominent, 



Frc. ^o— Lotus coriuc!,laliis{\iie\). 



and being joined together to the apex, and having the 

 petals folded over one another when not joined, affords a 

 broad and easy alighting place for insects. The tenth 

 stamen is separate at the base, and the staminal 

 tube so formed that it may contain nectar. Whether it 

 does so or not I have not observed. The filaments are 

 stiff, and the pollen sufficiently dry and dusty to come 

 out in abundance on pressure being applied to the folded 

 top of the keel. The stigma comes out first, and often 

 remains outside the keel, whilst the stamens, on pressure 

 being removed, resume their position. 



Trifoliiim repcns (Dutch Clover). — These flowers, 

 being in an umbel, afford a good foothold for bees, and 

 do not require an alighting place on each flower so much 

 as in the case of larger and separate blossoms. Never- 

 theless, they are upright in the bud, inclined in the flower, 

 pendent and recurved after blossoming. No flowers are 

 upright in full blossom, and consequently the centre or 

 summit of the umbel becomes bare. The flowers thus 

 tend to the usual position, even though in an umbel. 



The claws of the wings and keel are united, and form 

 a half tube, containing within them the staminal tube. 

 The tenth stamen is perfectly free, and the staminal tube, 

 as usual in such cases, contains nectar. Bees are fond of 

 the flowers, and must, in entering the half tubes of the 

 keel and wings, meet the stigma and carry away 

 pollen. 



Trifoliiim pratiusc. — The position of the flowers in 

 the umbel changes as in T. repcns, though in a less 

 marked manner. 



The long claws of all the petals, including thevexillum, 

 are united so as to form a complete lube, at the bottom of 

 which is much nectar. The limb of the keel is open at 

 the top, but the aperture is small, so that an insect enter- 

 ing cannot fail to touch both stigma and anthers. 



The filaments of the nine stamens cohere to one 

 another, and to the tube of the corolla from the point of 

 union of the petals, so that there is no separate staminal 

 tube. The tenth stamen is entirely separate lor its 

 whole length. Looking to the course of the apparent 

 veins of the petals and stamens on the tube, it seems as 

 if the vexiUum really formed the tube, and as if the nine 

 united filaments of the stamens by themselves would 



