PART III. | THE MECHANISMS OF FLOWERS. 485 
been quite wrongly ascribed to many other Labiates. The styles 
in these four genera are still applied closely together after the 
anthers have dehisced, and though when the styles separate the 
_ anthers may be still coated with pollen, self-fertilisation is pre- 
vented by the style projecting far beyond the anthers. In species 
where cross-fertilisation is less perfectly insured, and which pro- 
duce seed frequently by self-fertilisation, the smaller hermaphrodite 
_ flowers can never become transformed into purely female flowers, 
| since their anthers are always of real service, and cannot be 
_ removed by natural selection. And (2) it cannot be doubted that, 
in Nepeta, Thymus, Origanum, and Mentha, the size of the flowers 
has been subject to great variations, since the same condition 
occurs even now, at least in Nepeta and Thymus. While flowers 
of the same age on a single plant differ little in size, we find on 
_ different plants all intermediate grades in size, between the smallest 
_ female and the largest hermaphrodite flowers ; and the largest female 
flowers are quite as large as the smallest hermaphrodite. 
Darwin’ adduces several weighty objections to this view, and 
my own observations on Centaurea Jacea also militate against it. 
Darwin prefers to ascribe the gynodicecious condition to increased 
fertility. He supposes that either some individuals tended to pro- 
duce more seed, and consequently less and less pollen, until finally 
_ their stamens disappeared; or else that the stamens of certain 
| individuals began to show a tendency to abort, and that these 
individuals consequently produced more seed. 
‘ Visitors: A. Hymenoptera—Apide : (1) Bombus agrorim, F. § 2 (10— 
_ 15); (2) B. confusus, Schenck, 2 (12—14) ; (8) B. lapidarius, L. § Oe : 
(4) B. hortorum, L. § 2 (18—21); (5) B. pratorum, L. 2 (11—12) ; (6) B 
_ Rajellus, Ill. 9 (12—13) ; (7) B. silvarum, L. 9 (12—14) ; (8) B. (Apaths) 
i ~ Barbutellus, K. 9 (12); (9) B. (A,) vestalis, Foure. 9 (12); (10) B. (A.) 
il rupestris, F. 9 (11—14) ; most of these species very ab., sucking both (nine 
_ and hermaphrodite flowers in the normal way ; (11) B. terrestris, L. 9 (7—9), 
_ always bores through the corolla of the hermaphrodite flowers, and sometimes 
of the small female flowers, though its proboscis is long enough to suck the 
latter normally ;—sometimes it must at least attempt to suck the hermaphrodite 
flowers normally, for once (May 7, 1871) I saw a specimen sucking female 
flowers with pollen of Nepeta on its head ; (12) Apis mellifica, L. $, sucking 
the female flowers normally ; once (May 3, 1871) I saw it insert its head into 
' several hermaphrodite flowers, but afterwards take to piercing the corolla with 
| its maxille ; (13) Anthophora pilipes, F. 9 ¢ (19—21), very ab. ; (14) Osmia 
| wnea, L. 2 (9—10) scarce ; (15) O. rufa, L. 2 $ (7—9), sucking both kinds 
of flowers normally ; (16) O. fusca, Christ. 2 (8) ; (17) Nomada varia, Pz. ¢ 
1 Forms of Flowers, p. 304, ete. 
