SELF-FERTILIZATION OF PLANTS. 347 
for there are some flowers which one might, on à priori grounds, assume to be self- 
fertilizing, but are, nevertheless, proterandrous ; such, for example, is Lycopus europeus, 
with its stamens reduced to two. Hydrocotyle vulgaris, as far as concerned the spe- 
cimens I examined, was decidedly proterandrous ; nevertheless the flowers were com- 
pletely concealed beneath dense foliage of heath, grasses, &c., and not at all likely to be 
seen by insects; yet every fruit was “set”’*. Again, absence of scent is not an invariable 
rule ; for Salvia clandestina, though adapted for self-fertilization, is still stron gly perfumed, 
These cases compel caution in anticipating results which may prove not to occur. 
7. Cleistogamous flowers + are always self-fertile. 
Mr. Darwin experimented upon two plants, Vandellia nummularifolia and Ononis 
minutissima. The following is a brief summary of his results. With regard to the 
first, the ratio of the height (100) of intercrossed conspicuous flowers and of self-fertilized 
conspicuous, and of the same with intercrossed and self-fertilized cleistogamous flowers, 
are as follows:—100:99 and 100: 94 respectively, a second comparison of the ratios 
of the heights of the former pair being as 100:97. "These results are practically ratios 
of unity. The crossed plants were, however, inferior in fertility to the self-fertilized 
cleistogamous flowers. Six of the finest plants of each kind yielded respectively 598 
and 752 capsules, all being from cleistogamous flowers, as the crossed plants did not 
produce conspicuous flowers that year. The number of seeds per capsule was as 100 : 106. 
As with Vandellia so with Ononis, the cleistogamous flowers were highly self-fertile. 
The crossed and self-fertilized conspicuous flowers yielded seed in the ratio of 100 : 65 ; 
but the ratio between the crossed and cleistogamous was 100: 111, and the seeds them- 
selves looked finer even than those from the erossed perfect flowers. 
Further details on. certain cleistogamous flowers will be found in my descriptions of 
British self-fertilizing plants (postea, pp. 351, 359, 374), 
8. Special adaptations Sor securing self-fertilization, mainly of British plants. 
RANUNCULACER.—Ranunculus hydrocharis, Spenner. Mr. Hiern (Journ. of Bot. 
vol. ix. p. 45) has made an excellent analysis of the many forms of this species of Ranun- 
culus, and deduced geometrical representations of the twelve species given in Babington's 
‘Manual of British Botany. But it appears to me that another desirable investigation 
would be to test the raison d'étre of these different forms ; and I would suggest the fol- 
lowing interpretation as being more or less probable. Mr. Hiern finds that the twelve 
species can be grouped under three heads, and that if each species be placed ina given 
plane with reference to two axes of coordinates, the abscissa being the same number of 
units of length as the normal number of stamens, and the ordinate being the number of 
veins on each petal, then the species heterophyllus, confusus, Baudotii, trichophyllus, and 
Drouetii lie in a straight line whose equation is z—45--11—0; the species peltatus, flori- 
* Perhaps minute flies may assist, as I observed such diligently sucking the honey from Hydrocotyle americana in 
Kew Gardens. Indeed, Ido not deny that any flower, however minute, may not be visited, as Chickweed is sometimes, 
by bees (Meehan); but my object is toshow that it is quite immaterial, even if such be capable of self- fertilization. 
` * For lists see Darwin's * Forms of Flowers,’ p. 312, and Bot. Zeit. xxv. p. 67. See also some excellent remarks by 
Prof. Oliver in the * Natural History Review,’ 1862, p. 238 seqq. ^ 
