500 



NATURE 



\Oct. 17, 1872 



Ononis arvcnsis. — The vexillum in the opened flower is 

 perpendicular or a little bent back ; and the v/ings, which 

 are small, are also perpendicular, so that an insect may 

 li"ht either on the vexillum or on the wings, and has to 

 llirust itself belween the vexillum and the wings. 



The keel is long and pointed as in Lotus and Lupin : 

 ar.d the stamens push out the pollen as in those flowers. 



The stamens aie quite monadelphous, the staminal 

 tube is close fitting, and there is no nectar and no space 

 for nectar within it. The humble-bee certainly does not 

 put his proboscis down the tube, but between the tube and 

 the vexillum. 



Antliyllis Vulneraria. — This flower being in an umbel 

 needs no peculiar position to give insects a foothold. Its 

 pecuharity is that the large calyx, the sepals of which 

 cohere up to their narrow mouth, forms a dilated tube or 

 vessel which contains abundance of nectar. The limbs of 

 the wings are attached to the keel, but the claws of all the 

 petals are long, narrow, threadlike and perfectly free, so 

 as to leave free access to the nectar when the proboscis 

 of an insect has once passed the mouth of the flower. 



The aperture between the vexillum and the coherent 

 wing and keel is however very narrow, so that an insect 

 in passing it cannot fail to push the keel outwards and 

 bring out the stiff style and stamens. The filaments are 

 entirely joined together, and form along close fitting tube 

 in which there is space neither for nectar nor for the pro- 

 boscis of an insect. 



Ulcx nanus (Autumn Gorse.) — This flower is upright in 

 the bud, assumes the usual horizontal position when in 

 blossom, and reverts to the upright position in the pod. 

 The bud is protected by a stout, large and hairy calyx, 

 and the pod is stout and hairy. 



The wings arc perfectly free from the keel, and the 

 rounded lobes of the keel are separate from one another 

 at the extremity and for a considerable part of the lower 

 side, so as to make the flower comparatively open. The 

 stamens and pistil are stiff, and come out on the keel 

 being pressed down ; and the pollen, which is dusty, comes 

 out in a cloud. 



The staminal tube is perfectly closed and close fitting. 

 There is no cavity in it for nectar, and no aperture at the 

 base. There are traces of nectar on the veins and in the 

 hollows of the vexillum, especially on the midrib and in 

 the hollow towards the base. The bees are fond of it. 

 They settle on the keel and thrust their heads between it 

 and the vexillum, pushing the latter upwards. In 

 struggling to do this their legs are in violent motion on 

 the top of the keel, pressing it down. In doing this they 

 invariably open it, make the anthers project, and dust 

 their own body with pollen ready to meet the stigma of the 

 next flower (see Fig. 17). 



The contrast between the free wings, the obtuse and 

 semi-attached lobes of the keel, the stiff filaments, the 

 hairless style, and the dusty pollen of Ulex, and the 

 adherent wings, closed keel, moist pollen, and brush-clad 

 style of Pisum, Vicia, Lathyrus, &.C., and the correlation 

 between these differences, having regard to the ultimate 

 object in both cases, viz. the conveyance of pollen by an 

 insect, are very striking. 



U/cx Europaiis is similar to Ulcx nanus, and I have 

 seen bees upon it in April. But are there enough of these 

 insects abroad during the winter season, when this llo%ver 

 blossoms, to fertihse it ? 



Gi'u/s/d Anglica. — The wings are separate from the 

 style, the keel is straight and horizontal, but is re- 

 flexed after maturity, probably when it has once 

 been visited by an insect. The anthers have stiff fila- 

 ments and dusty pollen, all of vv'hich is shed when the 

 flower is once opened. 



The style is stiff, and coils back on the opening of 

 the keel, whilst the stigma is oblique. When so curved 

 back the stigma would touch an entering insect. I 

 have not ascertained where the nectar is in this flower, 



but probably not in the closed and close-fitting staminal 

 tube. 



Sarotliavunis (Broom.) — The keel is perfectly free from 

 the wings, is obtuse and closed when the flower first opens. 

 In this stage the style is bent against the keel in 

 such a way that its stigma (which is at the extremity) 

 is turned away from an entering insect. At a touch 

 the keel opens and falls down by a sort of hinge, 

 and docs not recover its position. The stiff stamens 

 shed their dusty pollen, as in Ulex and Genista. The 

 elastic style at the same time coils itself inwards towards 

 the vexillum. In a few moments it has curved back so 

 far as to complete one spiral coil, and bring the stigma 

 round so as to meet an insect subsequently entering. In 

 opening the flower with the finger or a pencil, the stigma 

 does not catch its own pollen, but after recoiling can 

 hardly fail to rub the next body which enters. 



The staminal tube is complete, and there is no space 

 for nectar or apertui'e into it. I have not ascertained 

 where the nectar is to be found, but not, I think, in the 

 thin, close-fitting staminal tube. 



Cytisus [qu. nigricans ?) — common in London" green- 

 houses. — The raceme is terminal ; the peduncle is nearly 

 horizontal. The pedicels are set on all round the peduncle, 

 but in blossoming are so bent upwards as to make the 

 vexillum of each flower nearly upright, and the keel and 

 wings nearly horizontal, i.e., so as to bring the flowers into 

 the normal position. 



The separation of the petals, the reflexion of the keel, the 

 closed staminal tube, and the dusty pollen, are the same 

 as in Sarothamnus and Genista. There is no nectar, and 

 no place for any in the staminal tube. 



The above details seem to point to some generalisations 

 concerning papilionaceous flowers. 



1. The position of the flowers in blossom, whatever their 

 other wants and habits, is such as to make them attractive 

 and convenient to insects. In general the showy vexillum 

 is upright, and the keel and wings horizontal. This is 

 effected in various ways : by the raising and straightening 

 of the stalk, as in Pisum and Lathyrus ; by the lowering 

 of it, as in Faba, Phaseolus, and Ulex ; or by giving the 

 pedicel a half twist, as in Robinia, Wistaria, and Labur- 

 num. 



2. The cohesion of the petals (which in this single 

 tribe is so various) is in each flower correlated to the 

 position of the nectary, the structure of the fertilising 

 apparatus, and the nature of the pollen. Thus in Ulex, 

 Genista, and Sarothamnus, the cohesion of the petals is 

 at a minimum, the wings do not adhere to the keel, and 

 the keel itself is comparatively open. Correlatively the 

 filaments are stiff and the pollen dusty, and the insect 

 gets freely dusted with it, without aid from any union of 

 the petals. In Pisum, Lathyrus, Vicia, Phaseolus, and 

 others, the wings not only serve as a landing place for 

 insects, but, being united to the keel, serve to pull it 

 down and force out the pollen. In Trifolium, the co- 

 herence of the petals is at a maximum, and produces a 

 complete long tube containing much nectar, and having 

 the organs of fertilisation in the access afforded by its 

 narrow mouth. In Anthyllis the claws of the petals are 

 so thin and so free from each other as to afford no recep- 

 tacle for nectar, whilst the staminal tube is closed and 

 tight fitting, but the want of a nectary is made up by the 

 cohesion and form of the calyx. The various degrees of 

 cohesion between the petals of the keel — from the com- 

 paratively free keel of Genistcre, through the prolonged 

 acute keels of Lupin and Lotus, and the oblique keel of 

 Lathyrus, to the spiral tube of Phaseolus — and the adap- 

 tation of each of these forms to its own stamens and pistil, 

 is no less remarkable. 



3. The degree to which the cohesion of the stamens is 

 carried (so remarkable a feature in this tribe) appears 

 to depend en the necessity for access to ncclar. In 



