parti] THE MECHANISMS OF FLOWERS. 85 
which are so completely in relation to the movements of the bee’s 
head. The stamens in the unripe condition are bent down, when 
they dehisce they stand upright in the path of the bee’s head, 
and when they are withered they sink completely down. The 
_ carpels erect themselves after the withering of the last of the stamens, 
_ and their stigmas then stand in the way of the bee’s head. Self- 
fertilisation cannot take place in absence of insects. Of the two 
native bees whose proboscides are of sufficient length to suck the 
honey of this flower thoroughly, Anthophora pilipes has finished its 
period of flight and has disappeared when D. elatum flowers. B. 
hortorum therefore remains as the only indigenous bee fitted for 
sucking up the honey of Delphiniwm, and it is in fact found in 
great abundance on the flower. Many of our other species of bees 
are well fitted for sucking a portion of the honey, but I have never 
seen any of them, or any other species of insect, sucking on D. clatum. 
At Strassburg, D. elatuwm is visited also by Anthophora personata, 
IIL. (590, 1). 
Delphinium Staphysagria, L., described and figured by Hilde- 
brand (356, p. 473) agrees in most points of its floral arrangement 
with D, elatwm, and is, likewise fertilised by humble-bees. 
15. DELPHINIUM Conso.ipA, L., differs in the arrangement of its 
flower from D. elatwm chiefly by the coalescence of the four petals 
_ into a single piece, which leads to the following modifications. The 
_ two upper petals unite by means of their backward directed processes 
into a single spur, the pointed end of which secretes and contains 
_ the honey ; their expanded portions, which are turned forward, also 
_ coalesce lengthwise with one another, and therefore cannot be thrust 
apart on the entrance of the bee’s head; but in conjunction 
with the lower petals they form a sheath which is well adapted to 
receive the bee’s head, and is only open below. | Here in the first 
_ stage of flowering it presents the anthers, and in the second the 
_ stigmas, for contact with the under surface of the bee’s head. The 
lower petals, which thus form the side walls of the sheath or tube, 
unite with the upper ones, but not with one another, and yield 
laterally when the bee’s head is thrust in. Path-finders and 
a lower boundary of the entrance to the spur are wanting 
here. 
Since the cycle of development and movement of stamens and 
anthers agrees with that in D. elatum, cross-fertilisation is in like 
manner insured on the occurrence of insect-visits; and similarly, 
if they fail to occur, self-fertilisation is impossible. 
