6o4 SECONDARY CHANGES. 



to the parenchyma, and which must be the cause of the peculiar curling of the 



band-like stems or branches of these plants. 



According to the known facts, which certainly require to be completed and 



tested, the following conditions, which in the strict sense only partly belong to this 



category, occur here. 



The young internodes, with leaves in two rows, are bluntly four-angled. In trans- 

 verse section the pith lies in the form of a cross, the arms of which are in the plane of 

 the two orthostichies of leaves, and in one at right angles to it. It is surrounded by 

 a normal woody ring (medullary sheath) consisting for the most part of relatively narrow 

 and thick-walled, radially arranged elements, it is therefore dense : it is itself thicker 

 between the arms of the cross-like pith than outside the latter, and therefore it is 

 roundish or bluntly octagonal in transverse section. On this zone which scarcely 

 attains a thickness of i™™, and sharply marked off from it, a softer wood is deposited by 

 continued growth in thickness: it consists of wide pitted vessels, narrow hard sclerenchy- 

 matous fibres, and thin-walled, apparently non-lignified, fascicular parenchyma; these 

 elements are variously distributed, according to the species, at least according to the 

 specimens investigated, the names of which were for the most part not accurately deter- 

 mined ; this need not here be described in detail. Numerous narrow medullary rays, and 

 here and there single very broad ones, traverse the strands of wood, and are distinguished 

 in the dry material by the brown contents of their cells. They are as usual placed with 

 their peripheral ends perpendicular to the surface, and on the flat portions of the stem 

 they are also curved near the middle in a curve convex towards the margins. The 

 body of wood thus arranged now grows in thickness, especially at the two sides alter- 

 nating with the orthostichies of leaves, so that the whole stem attains the form of a 

 ribbon-like plate rounded off at the two margins. Perpendicular to the surface of the 

 plate the increase of wood is generally of much less extent, so that over the pith the 

 whole thickness of the stem is not greater, but often even less than at the sides. 



The thickening is usually stronger on one surface than on the other, so that the pith 

 remains nearer the latter. Local inequalities of thickening lead in most, but not in all 

 specimens at hand, to the formation of unequally thick and irregular longitudinal ridges 

 and furrows. The cortex shows no generally remarkable peculiarities of structure, 

 it is on the average of equal thickness round the stem, if the numerous local irregularities 

 be not taken into account. 



Older ribbon-shaped branches of Caulotretus, about 40-50™™ broad and 10™™ thick, 

 show in their form the peculiarity that their margin is considerably shorter than the 

 middle, the latter is therefore strongly undulated, and often in a direction very regularly 

 perpendicular to the surface ; the undulation is strongest in the middle around the pith, 

 at the margin it is not present at all, and towards the latter it gradually diminishes '- 

 Each crest and each hollow of a wave, i. e. each point of strongest curvature towards the 

 two surfaces of the stem, corresponds to the insertion of a leaf, or axillary shoot ; the latter 

 is placed in the cases investigated somewhat higher (in acroscopic direction) than the 

 last-named point. Young shoots, on the other hand, even those which are already flatly 

 ribbon-shaped (e.g. one before me 10™" broad and 3-4™™ thick) show no undulations, or 

 have them hardly indicated at all ; as they grow older the undulations become steeper. 



According to these data the undulation must arise from the fact that with the 

 successive growth in thickness either the margin of the plate becomes absolutely 

 shorter, or the middle absolutely longer than at first. The former alternative is 

 a priori improbable, and is not made more probable even by any direct observation. 

 The other condition might, with reservation till exact quantitative measurements be 

 made, be explained partially on the assumption that, as in many normally growing stems 

 (p. 505), the elements of the successive zones of growth increase in length, but here to 



' Compare the drawing in Duchartre, Elem. de Botanlque, p. 166, Fig. 77. 



