514 MODIFIED CIECUMNTJTATION Chap. X. 



upright flo-wer-heads, arising from branches in all sorts of posi- 

 tions, on plants growing in a warm greenhouse, were niarked 

 with thread, and after 24 h. six of them were vertically depen- 

 dent; thfse therefore had travelled through 180° in this time. 

 Ten were extended sub-horizontally, and these had moved 

 through about 90°. Three very young peduncles bad as yet 

 moved only a little downwards, but after an additional 24 h. 

 were greatly inclined. 



At the time when the flower-heads reach the ground, the 

 younger imperfect flowers in the centre are still jsressed closely 

 together, and form a conical projection ; whereas the perfect and 

 imperfect flowers on the outside are upturned and closely sur- 

 round the peduncle. They are thus adapted to offer as little 

 resistance, as the case admits of, in penetrating the groundj 

 though the diameter of the flower-head is still considerable. 

 The means by which this penetration is effected will pre.sently 

 be described. The flower-heads are able to bury themselves in 

 common garden mould, and easily in sand or in fine sifted 

 cinders packed rather closely. The depth to which they pene- 

 trated, measured from the surface to the base of the head, was 

 between i and J inch, but in one case rather above 0'6 inch. 

 With a plant kept in the house, a head partly buried itself in 

 sand in 6 h. : after 3 days only the tips of the reflexed calyces 

 were visible, and after 6 days the whole had disappeared. But 

 with plants gi'owing out of doors we believe, from casual obser- 

 vations, that they bury themselves in a much shorter time. 



After the heads have buried themselves, the central aborted 

 flowers increase considerably in length and rigidity, and 

 become bleached. They gradually cm-ve, one after the other, 

 upwards or towards the peduncle, in the same manner as 

 did the perfect flowers at first. In thus moving, the long claws 

 on their summits carry with them some earth. Hence a flower- 

 head which has been buried for a sufficient time, forms a rather 

 large ball, consisting of the aborted flowers, separated from one 

 another by earth, and surrounding the little pods (the product 

 of the perfect flowers) which lie close round the upper part of 

 the peduncle. The calyces of the perfect and imperfect flowers 

 are clothed with dmple and multicellular hairs, which have the 

 power of absorption ; for when placed in a 4eak solution of 

 carbonate of ammonia (2 gr. to 1 oz. of water) their proto- 

 plasmic contents immediately became aggregated and afterwards 

 displayed the usual slow movements. This ciover generally 



