SELF-FERTILIZATION OF PLANTS. 331 
plant seeds abundantly as long as the weather is at all ** open." The rapidity in doing 
this, as with many other self-fertilizing plants, is very astonishing. Any flower assumed 
to be no. 1 will be shedding its ripe seed, while no. 3, in order of development, is a self- 
fertilizing bud. Spergula arvensis also seeds freely while in bud; but as soon as a warm 
sun-shining day comes in January the flowers will expand.  Cerastium glomeratum, 
however, I found produced nothing but closed self-fertilizing buds, even in an intensely 
and exceptionally hot day in June 1876. "The plants were growing, too, in the middle 
of an exposed hay-field. Polygonum Convolvulus seems never to open its flower-buds. 
Hence many flowers, though opening in warm sunshine, often fail to do so, but still 
set seed freely on other occasions; and in all such circumstances self-fertilization is the 
object. gained. 
4. vii. Flowers may become habitually self-fertilizing in the absence of insects specially 
adapted for intercrossing them. Although there is reason to suspect that when an 
annual or biennial, with dichogamous flowers, fails to receive the visits of insects, as 
when it is transferred to another country, it will become extinct, yet there are some cases 
where it appears, at least when under cultivation, that they may become independent of 
the visits of insects, and so propagate themselves by self-fertilization. 
The Papilionacez are specially constructed for facilitating the intercrossing of different 
flowers by the agency of insects. Mr. Darwin quotes the description of Mr. Farrer 
(1. c. p. 160) given in ‘ Nature,’ Oct. 10, 1872. But as all irregular flowers may be assumed 
to have been developed in adaptation to insects, I take this for granted. In England 
and N. Germany, however, the Garden Pea, Piswm sativum, is rarely, if ever, visited by 
insects; and Mr. Darwin observes, “It does not follow that the species in its native 
country would be thus circumstanced.” Lathyrus odoratus is invariably self-fertilizing 
in this country; and even when visited by tiumble-bees these insects do not appear able 
to depress the keel-petals sufficiently so as to expose the anthers and stigmas; but at 
Florence “it is the fixed opinion of gardeners there that the varieties [of the Sweet 
Pea] do intereross, and that they cannot be preserved pure unless they are sown 
separately." 
Phaseolus is another genus illustrating the same fact. P. multiflorus produced with 
. Mr. Darwin from 4 to 4 the number of pods when covered which were formed on 
uncovered plants. It is only morphologically sterile, for it can produce “remarkably 
fine pods " when the flowers are simply and mechanically moved. 
P. vulgaris, on the other hand, is * quite fertile " when insects are excluded. 
Mr. Darwin observes that *this difference in self-fertility between P. vulgaris and 
multiflorus is remarkable, as these two species are so closely related that Linnæus 
thought that they formed one.” 
Lastly, Lupinus luteus and L. pilosus seeded freely with Mr. Darwin when insects 
were excluded ; but the flowers of some species of Lupine will not do so in New Zealand 
unless artificially disturbed (see Z. c., note, p. 150). 
These, and such like cases, seem to indicate that certain plants, with conspicuous 
flowers, and specially adapted for intercrossing by insect agency, may, by transportation 
to foreign countries, and of cooler climate, where their particular insects are absent, 
