454 THE FERTILISATION OF FLOWERS. [parr m1. 
in a legitimate manner flowers which have been bitten through 
and robbed. 
The power of self-fertilisation has been completely lost. 
Visitors: Hymenoptera—Apide: (1) Anthophora retusa, L. ¢ (16—17) ; 
(2) Bombus hortorum, L. 2 (20—21) ; (3) B. agrorum, F. (12—15) ; (4) B. 
lapidarius, L. 9 (12—14); (5) B. silvarum, L. ?( 12—14) ; (6) B. Scrim- 
shiranus, L. 9 (10), all sucking normally, very ab.,—B. Scrimshiranus, L. § 
(9), bites through the corolla close above the calyx, and so reaches the honey ; 
(7) B. terrestris, L. 2 (7—9), very ab., only reaching the honey by biting,— 
once I saw it gnawing the upper lip, above the anthers, 
Pedicularis rostrata, L., P. verticillata, L., P. tuberosa, L., were 
seen by Ricca to be visited by humble bees (665). On the Alps, 
I have observed in addition to these three species, P. palustris, L., 
P. recutita, L., P. asplenifolia, Floerke, and P. foliosa, L. (609). In 
P. verticillata, the calyx is swollen, and the lower part of the 
corolla-tube is bent at right angles within the calyx; the honey 
is thus guarded from Bombus mastrucatus, which tries in vain to 
reach it, and which frequently bites through the tubes of P. tuberosa 
and P. foliosa, ete. 
The six species of Pedicularis which grow in Nova Zembla are 
all fertilised, in Delpino’s opinion, by Bombus terrestris, and so also 
is P. Kanei, which is found in lat. 79°, on the west coast of Green- 
land (172, 352). If this is really the case, all these species must 
have a shorter tube than P. szlvatica. 
334. RHINANTHUS CRISTA-GALLI, L.—lIn this plant, as in the 
foregoing, the dry smooth pollen is shed on the insect’s head, but 
the details of the mechanism are very different. Hach anther lies so 
close to its opposite neighbour, and they dehisce so widely on their 
opposed faces, that both together form one pollen-reservoir, closed 
in by matted hairs with which the contiguous edges of the anthers 
are provided (1, 2, Fig. 156). These pollen-reservoirs are borne 
on stiff filaments, of which the anterior pair are close together at 
their bases, and are beset with sharp points on their inner sides (1) 
so that a bee cannot insert its proboscis there; higher up, for the 
space of a few millimetres below the anthers, they are smooth and 
wide enough apart to admit the point of a bee’s proboscis easily. 
As soon, however, as the bee pushes its proboscis farther in, it 
forces the filaments apart, and tears asunder the two halves of the 
pollen-reservoir. The pollen falls out on to the bee’s proboscis, 
and is prevented from being scattered at the sides by the fringe of 
hairs on the lower edge of the anthers, This pollen-mechanism is 
