SELF-FERTILIZATION OF PLANTS. 355 
and so degraded into self-fertilizing forms which we now recognize as distinct species or 
even genera. Compare, e. g., the form Spergularia marina (proper), with large corolla 
and ten stamens, with the subsp. neglecta, having small petals, three or four stamens, 
and which is highly self-fertile. 
In support of this view, I may remark that while it often happens that the number of 
stamens and the quantity of pollen is much reduced in self-fertilizing flowers, or, on the 
contrary, the number of ovules may be reduced, compensated, however, by the pro- 
fusion of blossoms, yet the full number of stamens and a large quantity of pollen 
is not unfrequently retained, which therefore indicates the retention of a former 
condition, when a larger amount of pollen was required for insect agency. "Thus there 
is only one seed in Fumaria officinalis, but the quantity of pollen which becomes fixed 
to both stigmas by their pollen-tubes is very great. Similarly with Chenopodium, the 
five anthers are retained, but there is only one seed to each flower. Conversely, cleisto- 
gamous flowers of Viola canina and buds of Stellaria media produce many seeds ; but 
the anthers may be reduced to two. | 
The Caryophyllacez and its allied Orders are therefore very instructive as far as fer- 
tilization is concerned. Dianthus, which is very strongly proterandrous, may be placed 
at one end of the scale and Sagina apetala, or some incomplete member, at the other ; 
and between these extremes there are degrees of differentiation from the utter impos- 
sibility of securing self-fertilization and complete self-fertility. Yet since such an 
extreme form as Dianthus can become highly self-fertile in two or three generations, 
as Mr. Darwin has shown, we may be assured that all others could readily become self- 
fertilizing under fitting conditions. 
Plants of this Order also well illustrate the extreme rapidity of maturation of seed 
resulting from self-fertilization. If, for example, we call the terminal flower of a cyme 
of Cerastium triviale as No. 1, then if this be shedding ripe seed, the two capsules 
(No. 2) will be nearly ripe, while the four (No. 3) will be self-fertilizing buds; and lastly, 
the eight buds (No. 4) will be in a very rudimentary stage, with the corolla quite unde- 
veloped. The same rapidity may be noticed in the case of Poa annua, which is mostly, 
I suspect, self-fertilizing, and by the extraordinary rapidity of its seeding and growing 
so rapidly clothes bare places in our London parks, to the general surface of which, 
indeed, it contributes in no slight degree. This extreme rapidity of seeding com- 
pensates, where needed, for any comparative paucity in the number of seeds produced 
in a single capsule, though, as a rule, there is not even any particular deficiency in that 
respect at all. 
Other ‘points of importance are well illustrated by this Order, namely, the actually 
beneficial effects of size and duration of existence in these self-fertilizing plants. As 
a rule, they are small and annuals. Thus while Cerastium arvense and Stellaria - 
Holostea are perennial, the self-fertilizing Cerastia and Stellaria media are annuals. 
Similarly in other Orders, Geranium sanguineum, pratense, &c. are large-flowering and 
perennials, but the small-flowered species are annuals and more or less decidedly self- 
fertilizing. It is by no means invariably the case; e. g., Malva PONAM and mec Tn 
of ME are EENG and pique. 
