HETEROTAXY. 177 



supply of nutriment, and other means all based on the 

 same principle. Some of the Cape bulbs {Cyrtanthus) 

 are known not to produce their flowers till their leaves 

 have received, in some manner, a check. Fires which 

 often destroy the herbage thus have the effect of throw- 

 ing the plant into bloom. A very remarkable instance 

 is recorded of the production of flower-buds after an 

 injury to the leaf-buds in the ' Bulletin of the Botanical 

 Society of France,' vol. ix, p. 14G. It appears that 

 during the war of the French against the Arabs in 

 Algiers, the latter planted several hundreds of Agaves 

 with a view to obstruct the passage of the French 

 cavalry. The soldiers hacked these plants with their 

 sabres, and cut out the central tuft of leaves, or the 

 heart, as gardeners call it. The following season 

 almost every one of these Agaves sent up their large 

 handsome flower-spikes. It is well known that, under 

 ordinary circumstances, these plants do not flower 

 except at long intervals of time. 



Presence of flowers on spines. That the spine, as a con- 

 tracted branch, should occasionally produce flowers is 

 not to be wondered at, though the occurrence is by no 

 means common. M. BaiUon sli9wed at a meeting of 

 the Botanical Society of France (' Bulletin,' vol. v, 1858, 

 p. 316) a branched spine of Gleditschla bearing a flower 

 at the end of each of th^ subdivisions. ' This was, 

 therefore, strictly analogous with those cases in which 

 the peduncle is normally spiney. 



Formation of flower-bud on the petals. An instance of this, 

 it is believed, the only one on record, is cited in the 

 * Gardeners' Chronicle' for 1865, p. 769, by the Rev. 

 M. J. Berkeley, who describes the formation of a 

 flower-bud on the surface of a petal of Clarlda elcgans. 

 Reasoning from analogy there seems no reason why 

 buds should not be formed on the petals as well as on 

 the leaves. 



12 



