592 SECOND A Ry CHANGES. 



The structure of the individual zones of thickening has, with the exception of 

 some few cases to be rather more accurately described below, been very incompletely 

 studied, and further investigations are the more to be recommended, since very 

 numerous difierences appear in the special phenomena, at all events in different 

 species and genera. What follows can but give some indications of this. 



As has been already stated above, the structure of each zone of cambium and 

 the succession and direction of the divisions in it are in general, and with the 

 exception of individual characters still to be ascertained, the same as in normal 

 cambiums. The same holds for the originally radial direction, and the subsequent 

 displacements of the mother-cells of tissues or elements successively derived from 

 the cambial layers. For cases of extrafascicular cambium it may be added to these 

 statements in the first place that the latter is always derived, as far as investigated, 

 from tangential division of a single layer of parenchyma or meristem in contact with 

 the outer margins of the masses of primary phloem. The secondary thickening by 

 cambium begins while the differentiation and extension of the tissues surrounded by 

 it proceeds, and especially the leaf-trace bundles may continue to grow by formation of 

 nau ele?nents at the limit between xylem and phloem, after the manner of collateral 

 bundles (comp. p. 390). Longitudinal divisions of the cambial cells proceeding in cen- 

 trifugal succession add radially arranged secondary elements to the primary ones ; 

 most of them pass over directly, or at most after a single further longitudinal division, 

 into definitive tissue-elements, and are thus young wood-elements, or the direct 

 mother-cells of these. In this manner arise the elements of interfascicular tissue, 

 in many cases also those of the xylem of the bundles. But at certain points in the 

 ring of canibium there appear directly in one or a few tissue-mother-cells lying close 

 together, and separated off internally from the ring itself, rapidly repeated longitudinal 

 divisions facing in all directions; these form an initial strand (comp. Sect. 115) from 

 which, in the manner described generally for the collateral bundles, either a whole 

 vascular bundle is derived, or the phloem of a vascular bundle, the xylem of which 

 had already been begun by the simple tangential divisions in centrifugal succession. 

 In both cases the simple centrifugal tangential division proceeds further outside the 

 initial bundles ; outside each one of these interfascicular tissue is again formed, and 

 the same process repeats itself as the cambial ring is removed from the centre ; it is 

 repeated in the same transverse section successively on different radii, and in 

 successive transverse sections with such an arrangement of the initial bundles, that 

 the longitudinal course of the bundles, to be subsequently described, is brought about. 

 In the initial bundle, however, or the vascular bundle derived from it, the growth 

 is usually continued, when the increase at its outer margin has already ceased. 

 Tangential divisions are often found at the limiting surface between phloem and 

 xylem of a bundle, when the cambial ring is already separated from its outer 

 margin by several layers of cells in which division has ceased. 



In addition to the centrifugally formed fresh derivatives of the cambium there 

 are usually also centripetal formations, that is, a formation of bast : thus, e. g. in 

 slirubby Mesembryanthema, species of Obione, Halimus, &c. But in many cases, as 

 in the first internodes of the seedling of Chenopodium album, and Mirabilis Jalapa, 

 I have remained in doubt as to the appearance of centripetal divisions.^.. 



Wood and Oast consi.«;t, as has been stated, of collateral vascular buddies, and 



