6l6 SECONDARF CHANGES. 



All the vascular bundles above mentioned are, as far as investigated, fully formed 

 collateral bundles, which are not in contact with acti\e cambial layers ; also the 

 dissimilar tissue which surrounds them is fully developed. During the huge increase 

 in mass of the whole plant the structure remains the same in its chief characters, the 

 parts once present remain fundamentally unchanged/ only similar new ones are 

 added to them externally. This increase in thickness arises from a layer of cambium, 

 which runs close under the cortex round the whole periphery of the body : it shows 

 peculiarities to be mentioned below at the insertions of the leaves, and is least active 

 at the middle of the crown of old specimens. It consists of a few layers of radially 

 arranged cells, which are distinguished from the isodiametric cells of the adjoining 

 parenchyma by their arrangement as described, and their thinner walls, the latter con- 

 taining even here also the universally distributed granules of oxalate of lime ; their 

 radial is about half as great as their tangential diameter, or their height. Fresh young 

 divisions may always be observed in 1-3 adjoining cells of one radial row, the initial 

 layer in any case is thus only at most three layers of cells thick. The outermost 

 vascular bundles adjoin the cambial layer, others, apparently older, are separated 

 from it by layers of parenchyma, which are still radially arranged, but have begun to 

 be displaced. If the description in the case of the Chenopodiacese, Amarantacese, 

 and JNIirabilis on p. 592 be called to mind, and the phenomena to be described in 

 Chap. XVII in Monocotyledons, the data brought forward show that Welwitschia 

 has an extrafascicular cambium, forming at its inside vascular bundles which anas- 

 tomose in a radial and tangential direction, and alternate with interfascicular tissue. 

 The whole intra-cambial body is wood in the sense of the word given on p. 591. 

 Externally the cambium gives off a layer of bast consisting of parenchyma with 

 scattered sclerenchymatous fibres, again coinciding with the type of Mirabilis and its 

 allies. The production of this secondary cortex must be abundant ; the living layer of 

 cortex, which surrounds the cambium, is, it is true, but thin; for instance, in a trans- 

 verse section before me of the upper part of the tap root it is sixteen parenchymatous 

 cells in thickness ; thick crusts of bark, thrown off by repeated internal formation of 

 periderm, are superposed, as intimated above, on the older cortex. 



Towards the margin of the stem and the leaf-groove the cortex becomes 

 thinner, and the cambial layer less clear, since the whole tissue of the marginal 

 portion, as far as the inner surface of the leaf-groove and the insertion of the leaf, 

 with the exception of the thin outer cortical layer, consists of relatively delicate, regu- 

 larly seriate cells, apparently still growing and dividing, thus resembling those of the 

 cambium, between which lie sclerenchymatous fibres and small vascular bundles. On 

 this delicate, half-meristematic tissue, if the term be allowed, is inserted the 

 meristematic base of the leaf, which grows on in a basipetal direction. It may 

 accordingly be said that the extrafascicular cambium passes gradually over into the 

 half-meristematic tissue of the margin surrounding the leaf-groove : still it appeared 

 as though even here a cambial layer could be distinguished at a certain distance 

 from the surface. The doubt on this point must be dispelled by further investi- 

 gations. 



It is evident that the expansion of the margin throughout life is caused by the 

 growth of the half-meristematic mass of tissue of which it consists. It is also clear 

 that by means of it the bundles of the peripheral net running towards the margin 



