J 



u8 



SECOXDAKy CHANGES. 



develope into the bundles arranged round the leaf-traces. One, and perhaps some few 

 closely grouped cells of the canibial ring divide tangential!)' ; from the internal cell or 

 cells an initial strand is formed by longitudinal divisions facing in different directions, 

 vhile the outer cells continue the centrifugal tangential division. Those products of the 

 latter, which adjoin the initial strand external!)', assume at once the properties of 

 relatively wide parenchymatous cells, the initial strand is therefore separated by paren- 

 chyma from the cambial ring, which successively retreats outwards. 



Simultaneously with the fust formation of the initial strands, a centrifugally successive 

 development of parenchyma on the part of the cambium takes place between them. 

 Since this development continues slowly, the cambial ring, which retreats slowly 

 outwards, forms a number of new initial strands in the manner above described, at 

 successively alternating points at its periphery ; this centrifugally progressive activity 

 accordingly produces those irregular circles of vascular bundles, separated by delicate 

 and wide-celled parenchyma, whicli surround the ring of the leaf-traces. The number 



of these bundles varies greatly according to 

 the strength of the internodes ; in weak seed- 

 lings it is often hardly 8-10 (Fig. 236), in strong 

 flowering shoots 2-3 times that number. During 

 the process described, the growth in the direc- 

 tion of the transverse diameter continues in 

 the middle of the transverse section of the 

 internode. The vascular bundles increase the 

 nimiber and size of their elements in the usual 

 way — especially the luiited bundles of the 

 trace ( f) are usually to be distinguished at an 

 early stage by their considerable size ; "the cells 

 of the whole interfascicular parenchyma in- 

 crease throughout in width. The cells derived 

 directly from the cambium must therefore be 

 displaced from their original radial direction, 

 and the more so since the cambial divisions 

 apparently succeed one another relatively 

 slowly. Even the cells of the cambial ring 

 itself often show an irregularly distorted ar- 

 rangement. Finally, a stage is arrived at, in 

 wliich the extension of the inner parenchyma ceases, and simultaneously the cell divi- 

 sions in the cambium succeed more quickly and in greater number without changing their 

 succession. Thus a ring of thickening is now formed, which consists of relatively 

 narrow elements, arranged in regular radial rows ; it is composed, on the one hand, of a 

 fundamental mass of sclerotic elongated cells, on the other, of vascular bundles embedded 

 in this fundamental mass ; for the origin and arrangement of these, fundamentally the 

 same may be said (that is, if details be not taken into account) as above for Mesem- 

 bryanthemum. The old internode thus shows apparently medullary bundles within 

 a dense woody ring (Fig. 236). 



In the thin first internodes of the main stem of seedlings I first saw no bast paren- 

 chyma derived externally from the cambium. Later this process may begin, and in 

 strong plants this is always or as a rule the case. Similar, but simpler conditions occur, 

 according to Nageli's description, in Pisonia hirtella; also the more complicated 

 phenomena which Nageli describes in the internode of Boerhavia scandens and Bcngain- 

 I'illen spectabilh are to be connected with the above. 



The growth in thickness of the root of Mirabilis begins like that in normal Dicoty- 

 ledons. Corap. p. 473. The vascular strand is rarely tetrarch ', usually triarch ; the 



Fig. 236. — Mirabilis Jalapa. Transverse section tlirnti^h 

 the first epicotyledonary internode of a young specimen. 

 which is already several internodes in heij^ht {15) ; 7/ and m as 

 in Fig. 235, ne.xt to vi are the lateral bundles ; round the eight 

 bundles of the trace are nine apparently medullary bund'es 

 in the parenchyma, which is left white; further outwards is 

 the thick ring of wood ; c cambial-zone. 



^ Van Tieghem, Symmelrie de structure. I.e., compare p.' 473. 



