If the visits of bees are prevented, and if the flowers are not dashed by the wind against any object, the keel never opens, so that the stamens and pistil remain enclosed.  Plants thus protected yield very few pods in comparison with those produced by neighbouring uncovered bushes, and sometimes none at all.  I fertilised a few flowers on a plant growing almost in a state of nature with pollen from another plant close alongside, and the four crossed capsules contained on an average 9.2 seeds.  This large number no doubt was due to the bush being covered up, and thus not exhausted by producing many pods; for fifty pods gathered from an adjoining plant, the flowers of which had been fertilised by the bees, contained an average of only 7.14 seeds.  Ninety-three pods spontaneously self-fertilised on a large bush which had been covered up, but had been much agitated by the wind, contained an average of 2.93 seeds.  Ten of the finest of these ninety-three capsules yielded an average of 4.30 seeds, that is less than half the average number in the four artificially crossed capsules.  The ratio of 7.14 to 2.93, or as 100 to 41, is probably the fairest for the number of seeds per pod, yielded by naturally-crossed and spontaneously self-fertilised flowers.  The crossed seeds compared with an equal number of the spontaneously self-fertilised seeds were heavier, in the ratio of 100 to 88.  We thus see that besides the mechanical adaptations for cross-fertilisation, the flowers are much more productive with pollen from a distinct plant than with their own pollen.

Eight pairs of the above crossed and self-fertilised seeds, after they had germinated on sand, were planted (1867) on the opposite sides of two large pots.  When several of the seedlings were an inch and a half in height, there was no marked difference between the two lots.  But even at this early age the leaves of the self-fertilised seedlings were smaller and of not so bright a green as those of the crossed seedlings.  The pots were kept in the greenhouse, and as the plants on the following spring (1868) looked unhealthy and had grown but little, they were plunged, still in their pots, into the open ground.  The plants all suffered much from the sudden change, especially the self-fertilised, and two of the latter died.  The remainder were measured, and I give the measurements in Table 5/58, because I have not seen in any other species so great a difference between the crossed and self-fertilised seedlings at so early an age.

Table 5/58.  Sarothamnus scoparius (very young plants).

Heights of plants measured in inches.

Column 1:  Number (Name) of Pot.

Column 2:  Crossed Plants.

Column 3:  Self-fertilised Plants.

Pot 1 :  4 4/8 :  2 4/8. 
Pot 1 :  6 :  1 4/8. 
Pot 1 :  2 :  1.

Pot 2 :  2 :  1 4/8. 
Pot 2 :  2 4/8 :  1. 
Pot 2 :  0 4/8 :  0 4/8.