ANOMALOUS THICKENING IN DICOTYLEDONS AND GYMNOSPERMS. 581 



the large-celled pith, and are formed by longitudinal divisions facing in all directions, 

 soon ceases to grow (single rows of them may develope into sieve-tubes). The outer 

 layers develope into elongated, afterwards thick-walled parenchyma. When the normal 

 cambium has begun its growth at the outer side of the ring of wood, radial extension 

 and tangential division begins in a middle layer, and this process is extended, as in 

 the origin of a normal cambium, from each group laterally over an annular layer 

 running along the whole inner side of the ring of wood. This now forms, as 

 in a normal cambium but in a reversed direction, wood and bast, the former being 

 affixed to the zone of parenchyma which covers the inner surface of the ring, 

 and pushing the bast towards the pith. Both products of the inner cambium 

 resemble in their structure the secondary wood and bast of the normal outer ring, 

 and have, like them, both medullary rays and annual rings. THeir naturally very 

 limited growth compresses the originally broad pith more and more. It remains 

 to be investigated how long the process lasts, when it is terminated mechanically, 

 perhaps after the complete compression of the pith, and what relation it has to 

 the frequent splitting of old stems. 



l-IG. 230. 



Fig. 231. 



1-lG. -232. 



Figs. 230,231.— Transverse sections through stems of not exactly defined species of Seijania or Paullinia. Natural size. 

 From Schleiden.Grundz. ab outer rings; c the main ring of the compound woody body. The cortex is white in both figures, 

 tile xylem (in the strict sense) dotted according to the distribution of the vessels. Medullary rays obvious in Fig. 231, in 230 

 not obvious ; also the pith is not shown in the outer rings of the latter figure. 



Fig. 232. — Serjania caracasana. Transverse section through a j'oung internode just above the node (10). After Nageli. 

 M'ithin the ring of scierenchyma s are the main ring h and four outer rings ; to the left below is one outer ring opening into 

 the main ring. The dark prominences of the rings poinfing towards the pith are the primary bundles of the leaf-trace. 



Indications of a similar formation appear, according to Sanio\ to occur in 

 Rumex crispus, but further investigations are necessary on this point. 



Sect. i88. Several partial cambiums and rings of growth, lying side by 

 side in the same transverse section, appear in the most exquisite development in the 

 woody stems, so often described since Gaudichaud ^, of climbing Sapindacese of the 

 genera Serjania, Patdhm'a, and Thinouia. The transverse section of the stems in 

 question shows several separate rings of wood surrounded by a common cortex, and 

 in most of them a larger main ring occupying the middle, and several smaller outer 

 rings arranged in a circle round it (Figs. 230, 231) ; more rarely five, or in exceptional 



* Botfin. Zeitiing. 1865, p. 179. 



^ Recherches, &c. /. r. Tab. XVIII. — Compare further A. de Jussieii, Monogr. des Malpighi- 

 acees, I.e. — Schleiden, Griindziige (3 Aufl.) II, p. 166. — Treviranus, Botan. Zeitnng. 1847, p. 353. 

 — Criiger, Botan. Zeitung. 1851, p. 401. — Schacht, Lehrbuch, II, p. 58. — Netto, Comptes rendus, 

 torn. 57 (1863), p. 554, and Ann. sci. nat. 4 Ser. torn. 20, p. 166. — Nageli, Dickenwachsthum d. 

 Sapindac, compare p. 463. — Radlkofer, Atti del Congresso Botan'; ten. in Firenze, 1874, P- ^O' ^nd 

 Monographie d. Galtung Serjania, Miinchen, 1875. [Also, Radlkofer, Entstehnng d. secundiiren 

 Ilolzkorper im Stamme gew. Sapindaceen Naturforscher, Vers. z. Miinchen, 1877.] 



