ANOMALOUS THICKENING IN DICOTFLEDONS AND GFMNOSPERMS. 595 



parenchyma and hard fibrous elements form large masses,' which appear in transverse 

 section as alternating, irregular, concentric zones or islands ; this arrangement occurs 

 e. g. in Chenopodium hybridum, and, according to Falkenberg, in many ]\Iesem- 

 bryanthema. 



(c) The intermediate tissue formed at the beginning of the activity of the 

 cambium assumes in many cases an exclusively parenchymatous nature where the 

 cambium is extrafascicular ; its cells are more or less displaced from their radial 

 arrangement, by reason of the more continued strong extension of the masses of 

 tissue surrounded by them, and their own considerable growih in the transverse 

 direction. They are therefore arranged similarly to those of the primary pith, and 

 further they remain thin-walled like the latter and those of the primary medullary 

 rays, and therefore form with these two parts a connected mass of wide and delicate 

 cells. Later, when the transverse extension of the pith and the tissues near it ceases, 

 the pith-like intermediate tissue is succeeded by a dense tissue, with structure and 

 arrangement similar to (a). Since the latter together with the vascular bundles 

 enclosed in it correspond to the usual appearance of dense ' wood,' it alone appears 

 in transverse section to represent the ring of wood, while the wide-celled internal 

 tissues surrounded by it, including both primary and secondary bundles, appear one 

 and all to be medullary. Such apparently viedullary bundles are to be distinguished 

 from true medullary bundles, i. e. such as are within the primary ring of bundles 

 (comp. p. 248). Their occurrence appears to be not uncommon in the plants in 

 question, and is subject to two chief modifications. In the first place, a number of 

 layers of pith-like intermediate tissue are formed outside the primary ring of bundles, 

 but between the secondary bundles only dense tissue ; the primary ring alone 

 appears to be medullary in this case — and further is often very near to the dense 

 secondary ring. Thus, e. g. in Chenopodium album, Atriplex patula, Celosia, and 

 Achyranthes ; in species of Amarantus there are also true medullary bundles of the 

 leaf-trace. In other cases pith-like tissue appears also between the inner secondary 

 bundles, so that the transverse section shows in addition to the primary ring also 

 one or several, apparently medullary rings of bundles, as in the stem of IMirabiHs, 

 Oxybaphus, and also other Nyctagineas. 



All that is necessary has already been said above about the structure of the 

 bast in the forms with successively renewed annular zones, for here, if the expression 

 be allowed, each annular zone has its own layer of bast. For the other forms it 

 may be again stated that the secondary layer of bast, when present at all, consists, 

 as far as investigated, only of relatively few layers of radially arranged parenchyma, 

 in which crystal-containing sacs are not unfrequently scattered. Bast fibres are 

 found, scattered or forming a dense ring, only at the outer limit of the primary basl- 

 layer of the stem, and are absent even there in many species; in the secondary bast 

 they are found in none of the plants under consideration. 



To illustrate the above, two or three examples may be more thoroughly described, 

 though space does not here allow of an exhaustive description of the somewhat com- 

 plicated phenomena. 



I. Firstly, the shrubby Mesembryanthema. I have specially in view a form named as 

 M. virens Haw. Other species resemble this, in the main, according to the notes of 

 Rcgnault and Falkenber^, while in others again, as the annual INI. crystallinum, greater 



Q q 2 



