ANOMALOUS THICKENING IN DICOTYLEDONS AND GVMNOSPERMS. 583 



relatively to it, i. e. the former is turned towards the middle of the stem. In the 

 others this orientation is so inverted, that all the xylem-portions of one excluded 

 group face towards one common middle point situated within the group itself. A 

 bundle may belong successively, in its obliquely longitudinal course, to different 

 sides of one excluded group, and to the main ring, and must therefore undergo 

 successive rotations. 



These arrangements appear on the first differentiation of tissues ; they even hold 

 for initial bundles still surrounded by meristem. As the differentiation of tissues 

 proceeds, the main ring, together with the pith surrounded by it, assumes the above- 

 indicated normal properties. Each excluded group does the same ; the strand 

 of meristem lying between its opposite xylem-masses developes into parenchy- 

 matous pith ; at the limit of its xylem and phloem a normal cambium appears 

 all round. The corresponding narrow bands of meristem between the main ring 

 and the outer rings assume the structure of (chiefly parenchymatous) cortical tissue, 

 which is continuous into the general outer cortex. As may be concluded from what 

 has been said above, the arrangement and form of the rings within this general 

 sheath vary in successive transverse sections; in general, according to the rule 

 formulated by Nageli, an external ring, followed upwards and downwards, remains 

 intact and unaltered as far as the next node. Here it opens into the main ring, 

 whereby the primary bundles alter their mutual positions, and are continued on the 

 other side of the node as an outer ring or as a fold. (Comp. Fig. 232.) 



The histological composition of the mature parts shows, as far as the not very 

 thorough investigations extend, no fundamental differences from the normal siruclure 

 of wood and cortex. The presence of a many-layered ring of sclerenchyma at tlie 

 inner limit of the surrounding outer cortex (Fig. 232) is general. This is at first 

 completely closed, and is burst by the progressive dilatation, and split with the same 

 phenomena, as were noticed on p. 543 for Aristolochia Sipho. 



In the genus Thinouia in perennial stems there appear, in addition to the 

 phenomena described, cortical zones of refiewed growth, as in the Menispermese and 

 Leguminosse. According to Criiger's drawings ^, which may here be quoted, these 

 appear very irregularly arranged, sometimes as portions of concentric rings, some- 

 times as isolated bundles. If I understand Netto rightly, subsequently renewed 

 peripheral zones appear even in species with an originally simple ring of xylem. 



Radlkofer's divided xylems, which are peculiar to a definite group of the genus 

 Serjania, are distinguished from the described compound ones, as above indicated, 

 solely by this, that only five, rarely six to seven, peripheral partial rings are formed, 

 without the middle or main ring. Five (or six to seven) groups of bundles bulging 

 strongly outwards, which only anastomose at the nodes, are present from the first ; 

 each grows by means of a surrounding cambium into a ring, which is in itself 

 normal. 



According to Radlkofer's investigations most of the numerous genera of the Sapin- 

 daceae contain only species which do not climb, and show a normal stem structure. 

 Twining and climbing species occur in the genera : Cardiospermum with non-wcody 

 stem, and angular but normal xylem ; and Serjania, Paullmia, Urvillea, and Thinouia with 



1 Botan. Zeitung. 1851, Taf. VIII. 



