448 



LECTURE XXVII. 



points of flat structures, then only these two systems of cell-walls are present ; if, on 

 the other hand, the growing-point is hemispherical, or conical, or of some other 

 similar shape — i.e. not merely flat, but forming a solid body — then there exists still a 

 third system of cell-walls, viz. longitudinal walls running radially outwards from the 

 longitudinal axis of the growing-point. 



It will conduce to intelligibility, however, if we here again confine our 

 further considerations to a scheme constructed arbitrarily, but according to definite 

 rules. We may, in the first place, take simply the superficial view of a longitudinal 

 section through a growing-point. Confining our attention to Fig. 284, the outline 

 JEI^ of which represents the longitudinal section of a conical growing-point, we may 

 premise that this outline, which is often nearly realised in nature, possesses the form 

 of a parabola, and that the chambering of the space occupied by the embryonic 

 substance of the growing-point again takes place in such a manner that anticlinal 

 and p ericlinal wall s cut one another at right angles. With this premiss we can now 

 construct the network of cells in Fig 284 according to a well-known law of geometry. 



FIG. 284. 



Given that xx represents the axis, and jy the direction of the parameter, all the 

 periclines denoted by Pp form a group of confocal parabolas. Similarly, all the 

 anticlines Aa form a system of confocal parabolas having their focus and axis in 

 common with the preceding, but running in the opposite direction. Two such 

 systems of confocal parabolas cut one another everywhere at right-angles. 



We may now see whether a median longitudinal section through a dome-shaped 

 and approximately parabolic growing-point presents a net-work of cells essentially 

 agreeing with our geometrically constructed scheme ; and we at once find in the 

 growing-point of the Larch, for example (Fig. 285), the corresponding internal 

 structure, simply noting that in the figure the tw o prot uberances dd disturb somewhat 

 ^^^jyj^B^^JIJ'^ ^^^P,JiS^^- These are young "leaf-rudim?nfs'T5uaa!ng'off"^oml^^^ 

 growing-point. However, we at once recognise the two systems of anticlines and 

 periclines, the curvatures of which it can scarcely be doubted cut one another at 

 right angles as in our scheme above; or the anticlines are the orthogonal trajec- 

 tories of the periclines. As in our scheme, also, only a few periclines under the apex 

 ^ run round the common focus of all the parabolas ; the others as they come from 



