SELF-FERTILIZATION OF PLANTS. 327 
creased size of the female corollas in the foregoing cases is due to a tendency to abortion 
spreading from the stamens to the petals ” (* Form of Flowers,’ p. 308). 
On the other hand, the Corymbiferze of the Composite furnish exceptions ; for the ray- 
florets are female, and have larger (ligulate) corollas than have the disk-florets ; while 
Centaurea and Viburnum Opulus carry us further; for the corollas are relatively even 
still larger, the pistil being arrested as well as the stamens. 
4. iii. The partial or total arrest of development of the corolla. The tendency to arrest of 
the corolla in many flowers, which, however, set seed very freely, has often been observed. 
Thus M. Boisduval says* that ** Viola palustris is often apetalous in the mountains, 
notably at Lautaret. In the environs of Paris this species has petalous flowers, which 
fructify very well [probably by insect agency], but never apetalous. It is the same 
with V. biflora, an essentially Alpine plant." He adds that E Zuppi a species of 
Northern Italy, never bears * with us" conspicuous flowers, which does not hinder it 
from fructifying. He also cites Prismatocarpus and some Campanulas, whose flowers 
have no corolla, which fact does not hinder them from fructifying. 
Mr. Berkeley also alludes to the want of a corolla being no hindrance to the setting 
of seed (* Gard. Chron.’ 1855, p. 36). 
“ Varieties occur, as, for instance, in the Sweet William, where not a trace of petal exists, and yet 
every ovary is impregnated ; but the contrary condition is not infrequent. In several species of Violet 
the early petaliferous flowers are often barren, while those which appear later in the season are pro- 
ductive [i. e. the cleistogamous apetalous flower-buds]. In Ajuga Iva fertile flowers occur indifferently 
with or without a corolla. In Ononis minutissima, in contradistinction to the Violets, the earlier flowers 
are apetalous, and equally fertile with the petaliferous flowers which are produced later in the season; 
and the same equal fertility belongs to both kinds of flowers in Helianthemum, in which genus, Lespedeza, 
&c.they are generally simultaneous. In some of these cases the stamens seem very imperfect.... A 
very limited number of pollen-grains may sometimes be found." 
The general conclusion I would draw from all such cases where the corolla is more 
or less reduced or entirely absent is, that the flowers are readapting themselves to self- 
fertilization ; the conspicuous corolla being no longer required for insect agency, the 
energy usually expended in developing it is now diverted into the pistil, while a not 
unusual degeneracy or “ contabescence” of some of the stamens accompanies the 
process, involving a reduction of the quantity of pollen, which I regard, not as any 
sign of “ injuriousness," but as a form of less expenditure of wasted energy. 
4: iv. A partial or entire loss of colour from the corolla is often correlated with self- 
fertilization. Generalizing from several observations, I came to the conclusion that what- 
ever cause may be at work to abstract or withhold energy from, it may be, the staminal 
whorl, together with the corolla and adjacent parts, tends to promote self-fertilization ; 
so that asthe absence of colour may be regarded as indicating a poverty of nutriment 
in the corolla, whenever some colour is normally present in the species, and of which 
white or pale-coloured forms are varieties, such would seem to be, if not a proximate 
cause, at least a concomitant of self-fertility. Several instances are recorded by Mr. 
| * Bull. Soc. Bot. de Fr. iii. p. 569. 
i 222 
