parrut] THE MECHANISMS OF FLOWERS. 209 
- membranous outgrowth (n, 4), and that the ale and carina are 
connected very firmly with one another and with the staminal 
- column. The union between the ale and carina is formed in the 
following way:—The two folds or grooves, which in the erect 
extremity of the carina separate the two pouches from the free 
margins, are continued along the whole horizontal part of the carina 
‘near its upper border, and where the erect part joins the horizontal 
(at a’, 4, 5, 6) they each form a broad, rounded pit in which a fold 
of the ala (a, 1, 2, 3) lies; further towards the base of the flower 
_ they form a deep, narrow depression on each side of the cleft of the 
carina (0’, 4), in which a process directed downwards and forwards 
from the upper margin of the ala (0’, 2, 3) fits firmly. This inter- 
locking is rendered more secure by the presence of numerous blunt 
projections on the marginal process of the ala (b), and in the full- 
- grown flower it is difficult to separate the ala from the carina 
without tearing. 
_ The union of the alse and carina with the staminal column is 
effected, as in Medicago sativa, by two long, pouched processes of the 
ale (directed backwards) which lie upon the column (d, 2, 3), meet- 
Ing one another in the middle line. When the al and carina are 
_ depressed, these processes clasp the sides of the column, and as 
soon as the pressure is removed they return to their former place 
on its upper surface. The vexillum closes down firmly on the 
_ alg, preventing useless insects, eg. flies, from obtaining the honey 
; which is secreted abundantly in the ordinary position and is 
reached by two unusually large openings. The vexillum, at the 
junction of its claw and lamina, bears two deep, narrow depres- 
sions which project below as two sharp ridges converging 
anteriorly (0, 1), and these fit firmly into two depressions in the 
alze (c, 2, 3). Though the stigma is at first surrounded by the 
pollen of its own flower, it is probable that in case of insect- 
Visits, cross-fertilisation is ensured. For this pollen with which 
the stigma is covered, is easily rubbed off if we draw the 
Stigma over a glass plate; but immediately afterwards the stig- 
matic papille get rubbed away and leave a line of clear sticky 
fluid on the glass. Owing to the force which the mechanism 
of the flower renders necessary, the stigma must be rubbed hard 
on the under side of the bee, and thus without doubt, as Delpino 
| supposes, its own pollen is rubbed off, its surface is rendered 
sticky by breaking down of the oS and new pollen is made 
_ to adhere to it. 
Whether in absence ‘of insects. the stigma finally becomes 
P 
Se 
