REJUVENESCENCE IN NATURE. 57 



preceding leaves, in this way leaving the old bulb and 

 descending deeper into the bosom of the earth.* 



A retrogression from the hypsophyllary formation to 

 the euphyllary, or even to the cataphyllary formation, is 

 far more rare than the periodical sinking from euphyllary 

 to cataphyllary formation. Ananas affords a normal and 

 universally known example of this case, the summit of 

 the inflorescence returning to the euphyllary formation, 

 attaining complete Rejuvenescence in the " crown," as it 

 is called, and when this is removed and planted producing 

 new blossom and fruit in the third year. The same phe- 

 nomenon is exhibited by the New Holland genera of 

 Myrtacca, Melaleuca, Callistemon, Beaufortia, and Calo- 

 thamnus, the brush-like spikes of which owe their strange 

 "growing-through," or innovation, to a similar recur- 

 rence to the formation of an euphyllary shoot from the 

 end of the hypsophyllary region. What are called the 

 viviparous grasses, <?. g. Poa bulbosa and P. alpina, which 

 occur only in this condition in many places, and behave 

 like Ananas, might appear as paradoxes, but here it is 

 really the hypsophyllary region which is made into an 

 euphyllary region by a retrogression ; t and the tufts of 

 euphyllary leaves arising in this way subsequently become 

 detached, to recommence the ascending development as 

 independent stocks. At the same time the behaviour of 

 these grasses is not that natural to the species, but that 

 of a monstrosity become a variety. Leafy shoots occur 

 not unfrequently, as a mere accidental monstrosity, at 

 the summits of inflorescences: I have seen them especially 

 fine in Plantago lanceolata, where the leafy shoot at the 

 end of the spike became developed into a new perfect 

 stock. Even in flowers, retrogression of this kind occurs 

 as a malformation ; well-known garden examples are 



* This is not the place for a minute description of this strange phenome- 

 non. The description of it given b^ Henry, ' Nov. Act. Cur.,' vol. xxi, p. 1, 

 leaves some questions still open, which I shall take up at another opportunity. 



t The lowest glumes of the spikelets are mostly unaltered here, many of 

 them even having flowers in their axils. Vide Mohl, 'Bot. Zeit.,' 1845, 

 p. 33, t, i, fig. 2. 



