48 GROWTH AND CELL-DIVISION 



thallus of Melobesia^ for the apical cells in each cell-row divide longi- 

 tudinally at fairly regular intervals, and so reduce the tangentially 

 enlarging terminal cells to their original size 1 . Similar relationships are 

 shown in sections across the secondary wood of young stems of Finns. 



When the rows of cells are curved so as to form a series of confocal 

 parabolas, we have an arrangement like that shown in the darker part of 

 Fig. 12. To obtain a close resemblance we must assume that each row 

 of cells broadens away from the apex and divides longitudinally at 

 regular intervals as in Melobesia. The arrangement of the dividing- 

 walls in the apices of roots and stems does actually correspond roughly 

 to a system of confocal paraboloids of revolution with trajectories 



FIG. 12. Geometric diagram of a root-apex (after 

 Sachs). AA = axis of revolution; K K K- periclinal ; 

 A A, />P= anticlinals. 



FIG. 13. Geometric diagram of the annual rings and 

 medullary rajs in an excentric wood-cylinder. The 

 construction is based on a system of excentric circles 

 whose centres lie on the axis of symmetry N S at 2, 3, 

 4. and 5. r = geometric trajectories ; v = the slightly 

 diverging actual course of the medullary rays. 



intersecting them at right angles. Sachs termed walls parallel to the 

 periphery, periclinals (KK, Fig. 12), and those corresponding to the ortho- 

 gonal trajectories, anticlinals (A A, PP, Fig. 12). The walls lying in the 

 plane of a longitudinal section are at right angles to the visible primary 

 anticlinals, and in a transverse section across the apex they will appear as 

 orthogonal trajectories of the periclinals, that is, as division-walls at right 

 angles to the surface. The series of walls in a particular transverse section 

 have been termed ' transversals ' by Sachs, and they will usually be formed 

 by the interceptions of the radii and circumferences of a system of 

 concentric or excentric circles. 



Even when the periclinals form circles or ellipses instead of parabolas, 



1 [A still more regular arrangement is shown by the thallus of Coleochaete scidata, as to the 

 origin of which cf. GoebeFs Outlines, Clar. Press, 1887, p. 47.] 



