OF TENDRILS. 269 



are prolonged into slender filaments more or less ana- 

 logous to tendrils ; thus, the glands of the petiole of 

 Passiflora ligularis are lengthened into long and nearly 

 cirrhiform filaments. 



'The petiolary tendrils of Smilax are difficult to be 

 understood with regard to their anatomical origin. 

 These climbing shrubs generally have the petiole 

 dilated at the base into a kind of sheath, which might 

 readily be taken for a stipule adhering to the petiole, if 

 the analogy with other Monocotyledons were not opposed 

 to it ; above this sheath there proceed from the petiole 

 two opposite, simple, filiform, or elongated tendrils ; are 

 they prolonged petiolary glands, as in Passiflora ligu- 

 laris ? Their position would appear to make us believe 

 it ; but as no Smilax bears petiolary glands, this hypo- 

 thesis is not admissible. Are they prickles prolonged 

 into a filiform appearance? The irregularity of the 

 position of the prickles of Smilax, compared with the 

 regular position of the tendrils, ought to clear up this 

 opinion. Are they prolongations of the petiolary 

 sheath ? This idea is founded upon the circumstance 

 that they arise from the extremity of it, and it is con- 

 firmed by the fact that in *S. herbacea, which has no 

 sheath, there are no tendrils. Lastly, the opinion 

 which appears to me to be the most likely is, that the 

 leaf has originally three segments, and that the tendrils 

 represent the two lateral ones, which are abortive or 

 transformed. When very young leaves of Smilax aspera 

 are examined, it is very difficult not to admit this opi- 

 nion ; it is confirmed by the analogy of the Smilaceas 

 with the Aroideae, which often have leaves with several 

 segments. 



Pedunculary tendrils are more frequent in nature, 

 and clearer as to their origin, than the preceding ; they 

 are, as the name is intended to indicate, produced by 



