344 REV. GEORGE HENSLOW ON THE 
and that the necessity for intercrossing, as far as propagation by seed is alone ĉon- i 
cerned, has arisen by insect agency, and that whenever or wherever the proper insects 
fail, one of the following results may happen :—(1) The plant will die out, especially if it be - 
an annual. (2) If a perennial, it may propagate itself vegetatively, and so maintain its | 
existence. (3) It may become readapted to other insects. (4) It may revert to self- : 
fertilization, (a) with a retention of a conspicuous corolla, as Pisum sativum, or (b) with a 
dwarfing of it, as in the case of Fumaria and Trifolium, or (e) lose it ge, or (a 1 
become absolutely cleistogamous” — 
I strongly suspect No. 1 or No. 4 (b) to be usually the case with seedlings fe: ` 
to distant countries, as will be more particularly adverted to hereafter. Prof. Dyer, in 
his review of Mr. Darwin's work, ‘Cross and Self-fertilisation, in * Nature, thinks 
that cleistogamous flowers represent the primitive condition. I take an opposite view, | 
and regard them, as well as inconspicuous flowers, to be in all cases (excepting gymnda ] 
sperms) degraded conditions, and for the following reasons :— 3 
(a) Had such features as are borne by cleistogamous flowers been characteristic of. 
primordial conditions, they would have most probably been correlated with embryonic | 
or primitive states of other organs; or, at least, such plants would be of a relatively low | 
grade. They are, however, found to exist in widely different orders, which are amongst 
the most highly differentiated, and in no other respect showing any thing whatever of a 
primitive character about them. 
(b) Every degree of degradation can be found between the normal Bowtie and the 
cleistogamous on the same plant. Thus in strong cultivated plants of Viola odorata, 
comparatively large cleistogamous buds occur with five petals, the posterior spurred, and 
the stamens with their nectariferous appendages, but completely self-fertilizing, while on 
wild specimens these flowers may be entirely apetalous. (Tab. XLIV. figs. 4a, b, and 
5 a, b.) 
(c) Many plants have flowers in an intermediate condition, either opening when th 
weather is warm and favourable, but remaining closed if it be inclement, as do many of : 
the Alsinee, Illecebracee, Polygonum, Ze, P. Convolvulus and Hydropiper even appear 
to be already perfectly cleistogamous, though P. aviculare still opens its minute honeyless ` 
blossoms. Similarly, Cerastium glomeratum sometimes never opens its buds, at other 
times, and indeed scarcely ever, is more than half-open. Fumaria officinalis and. some 
small-flowered clovers, which are self-fertilizing, might almost be ealled permanently 
cleistogamous ; for, as in their more conspicuous allies, which are intercrossed, the sexual 
organs are completely concealed, but in their case do not require insect agency. As 
such flowers, though extremely inconspicuous, have yet irregular corollas, they are clearly 
derived from some ancestral form, from which their conspicuous allies have also been 
descended ; and it becomes therefore a moral conviction that they are ee con- 
ditions from more highly conspicuous forms. 3 
I do not think the Amentifere form any exception to this rule; for in many there is 
a regular calyx, with the stamens opposite the sepals, which seem to indicate that they 
are degraded conditions with the corolla gone. Thus, if we take Ulmus as our starting 
point, we have a abr five-lobed itr and fixe: stamens suet, the s a 
* 
