434 THE FERTILISATION OF FLOWERS. [parr Lil, 
and the flowers would be less diligently visited by the humble- 
bees. 
Dr. Ogle found that Antirrhinum produced no seed when 
protected from insects by a tent of gauze (632). 
Maurandia, Ort., Chelone, L., and Pentstemon, L’ Her., are, ac- 
cording to Delpino, proterandrous, the style coming to occupy the 
place of the stamens. In the two last genera, the stamens are 
broad, fleshy, and hollowed out at the base to secrete and lodge the 
honey. Pentstemon is fertilised by Bombus, Anthidiwm, and Apis. 
(178, 360). The position of the fifth stamen in Pentstemon and 
its frequent partial or complete abortion has been discussed by 
Dr. Ogle (632). A very thorough account of the floral mechanism 
of Pentstemon is given by Errera (230). 
322. SCROPHULARIA NODOSA, L.—In most Lamiales and Per- 
sonales the anthers lie in two pairs, one behind the other, and 
Fria. 146.—Scrophularia nodosa, L. 
t.—Flower, in first stage, from the front (x 7). 
2.—Ditto, from below (x 34). 
8.—Older flower, fertilising itself, viewed from the side. 
a, calyx; b, corolla; ¢c, stamen; c’, fifth metamorphosed stamen; d, ovary; e, style; f, stigma 
g, nectary ; h, drop of honey. 
: 4—7.—Fifth stamen in various stages of reversion towards its original form (x 12). ° 
i, black membranous lobe ; k, anther-lobe. 
touch an insect-visitor on the dorsal surface ; the stigma then, to in- 
sure cross-fertilisation, must also touch the insect on the back, and 
the style can scarcely lie elsewhere than between the two pairs of 
stamens, along the upper part of the corolla. The superior stamen 
comes in the way, and accordingly disappears ;! it has no chance of 
reappearing permanently, for it is directly injurious, and is weeded 
out by natural selection. I have only once found a flower of 
Cf. Dr. Ogle, No. 632, p. 51. 
