M. Trecul on the Structure of the Leaves of Palms. 29 



was about 1^^ millimetre in length. Its two rolls (or series of 

 leaflets), a little less advanced than those of the preceding leaf, 

 were comparatively more divergent at the base. From the sheath 

 of this leaf issued the apex of a still younger one not more than 

 f millimetre in length. Its lateral rolls were only marked with 

 faint strise or transverse depressions towards the middle, and its 

 sheath also gave exit to the tip of another leaf. This latter was 

 not moi'e than \ millimetre in length ; its sheath, which was 

 short and thick, had a broad rounded opening about the middle 

 of the leaf, through which the naked apex of the stalk could be 

 seen. This sheath was surmounted by the nascent rachis, but 

 this presented no trace of leaflets. It was broad and depressed 

 in its median portion ; and thei'e was on each side a longitudinal 

 swelling of so little prominence that it required considerable at- 

 tention to perceive it at all. ■ These swellings are the origin of 

 the two rows of leaflets. 



Thus a leaf of Chamcedorea Martiana commences with a simple 

 circular cushion at the apex of the stalk. This cushion or rudi- 

 mentary sheath is produced obliquely into a prominence, which 

 is depressed on its inner surface. This becoming elongated into 

 a cone produces a longitudinal roll on each of its margins. These 

 two rolls or cushions, which are more inflated near the sheath, 

 where however they terminate in a short point, become more and 

 more contracted towards the apex of the rachis. Originally they 

 are smooth, but during their growth scarcely sensible undula- 

 tions are produced on each side of them (first on the inner side) ; 

 of these the first appear a little way from the base of each roll, 

 and they afterwards increase in number and attain the base and 

 apex of the rachis. 



Whilst all the parts of the leaf continue groM'ing, these un- 

 dulations, increasing in depth, become furrows which penetrate 

 by degrees into the interior of the roll, at length arriving at the 

 opposite side on the outer surface so as to produce a rupture; 

 but the furrows which penetrate from the outer surface towards 

 the inner cease advancing before reaching the latter, so that 

 scission takes place only at the sides of the outer surface. In 

 this manner are produced as many leaflets plaited in the direc- 

 tion of their median nervure as thei'e were ribs on the inner sur- 

 face ; but the separation of the leaflets is not completed in this 

 manner through their entii'e length; it stops near the apex, 

 which remains united to the side of the leaflet placed above it. 

 When the leaf emerges from its sheath and the leaflet expands, 

 this point of attachment becomes broken and the apices of the 

 leaflets are set free. The union of the leaflets is not the same 

 throughout the palms ; in Phcenix sylvestris, Acrocomia sclero- 

 carpa, &c., the points of the pinnules are attached to a cellulo- 



