.322 THE CROSSING OF FLOWERS. 



capitulxim are of unequal length. The marginal florets are rather shorter than the 

 central ones, so that the stylar branches of the former are lower than those of the 

 latter. But this is not enough to bring the pollen which has fallen from the higher 

 stylar branches on to the stigmatic tissue of these older lower ones — since the lower 

 are situated rather nearer the circumference of the capitulum, and it is therefore 

 necessary that the pollen-bearing styles should incline outwards if their pollen is 

 to reach its proper destination. This is what actually happens. The originally 

 straight and erect styles bend outwards at an angle of 70-90°, even before their 

 branches have separated, and while they yet retain the pollen which they have 

 ■collected from the anther-tubes. When it is thrown off, it thus unavoidably reaches 

 the lower stigmas of the older flowers. Or sometimes it happens that the divergent 

 stylar branches of the younger flowers with attached pollen come into direct con- 

 tact with the stylar branches of older flowers, and that geitonogamy is effected in 

 this way. 



Numerous other Composites whose capitula are composed entirely of tubular 

 hermaphrodite flowers exhibit the same processes as Homogyne, which has been 

 ■chosen here as a type. The Wormwoods of mountain heights, e.g. Artemisia Mutel- 

 lina and spicata, exhibit a slight deviation. In them the central florets are raised 

 above the marginal ones, not only by their greater length, but because the receptacle 

 on which they stand is considerably arched. Obviously the florets at the top of the 

 dome will stand higher than those round its circumference. In very many Com- 

 posites {e.g. in Boronicum glaciale and scorpioides, in Senecio cordatus, in Telekia, 

 Buphthal'mum, Anthemis, and Matricaria), the receptacle is at first flat or but 

 slightly arched; but during the flowering period it rises up so much that it assumes 

 the form of a hemisphere, or even of a cone. This elevation in Doronicum-caT^itala, 

 for example, amounts to 1 cm., and it is relatively even more in species of Anthemis 

 ^nd Matricaria. The immediate consequence of this change in the receptacle is of 

 -course an alteration in the direction of the flowers which stand on it. Flowers 

 which stood erect on the receptacle of the capitulum when it first opened, assume 

 later an almost horizontal position. But the most remarkable thing is that these 

 changes keep pace with the advancing development of the flowers. In capitulate 

 inflorescences the marginal flowers open first, and those in the centre last (see vol. i. 

 p. 740). The flowers of each outer series are therefore always further advanced than 

 those of the adjoining inner series, and when the mature stigmas are already opened 

 in the outer flowers, the pollen of the inner ones is only just being pushed out of the 

 anther-tubes and shaken off the stylar branches as they separate. A wonderful 

 •contrivance now meets our gaze — the mature stigmas of the outer flowers are 

 brought directly under the inner flowers so as to catch their pollen as it falls, a 

 condition brought about by the alteration in form of the receptacle on which all the 

 flowers stand. Sometimes the pollen does not need to fall, for the flowers stand so 

 closely side by side and above one another that the divergent stigmas of the older 

 flowers come at once into direct contact with the pollen of the younger flowers. This 

 is the case in various species of the Groundsel genus (Senecio) where the two stylar 



