444 



LECTURE XXVII. 



elliptical disc : one of the preceding transverse walls T is still present as a straight 

 transverse wall JiJi, and the series- of longitudinal walls referred to above also still 

 exists as a straight longitudinal wall in the direction of the major axis of the ellipse. 

 In accordance with the mode of growth mentioned, however, the other walls pre- 

 viously marked T have now become converted into hyperbolically curved anticlines A, 

 This figure is only constructed hypothetically, and agrees with the requirement that 

 a filament consisting of two longitudinal rows of cells shall become converted into 

 a stalked ellipse, under such conditions that the walls still cut one another at right 



angles. 



In Fig. 280, which represents the development of a gemma of Mar chan/ia, there 

 will be perceived at once, in Figs. I-IV, alterations of the cell-network which agree 

 in all essential points with the processes illustrated in the diagram (Fig. 279). ' ; 



On the other hand, however, it also happens that networks of cell-walls undergo' 



FIG. 280. Development of the ffemma of ^«rc/!rt«^«T. 



displacements by growth, of such a kind that the usually rectangular cutting of the 

 anticlines and periclines passes over into one more or Ifess oblique. This is occasion- 

 ally the case in growing-points, the growth of which we shall only consider later on, 

 It is to be recognised more easily and frequently, however, on transverse sections of 

 woody bodies with excentric layers. This process also may be made clear most easily 

 by means of a hypothetical construction. In Fig. 281 I have assumed that a mass of 

 wood which is circular, in transverse section possesses a pith situated very excentrically, 

 The first ring of wood, /, is equally thick all round ; but all the succeeding annual 

 rings, //- VII, have grown much thitker oii the north side (N) than on the south 

 side {S). In order to simplify the construction it is assumed that the circumference of 

 each annual ring is nevertheless circular. The scheme may be so imagined that the 

 points 2, 3 — 7 on the line NS are the centres of the six annual rings following one 



