Jan. 1900.] BOTANICAL SOCIETY OF EDINBURGH 347 



of the extracambial cells of one or more medullary rays, 

 those implicated being generally situated near one or other 

 margin of the bundle, but still within it, though in rare 

 cases this relation is somewhat difficult to determine, owing 

 to the secondary growth of the stem in thickness. Of 

 these rays, one, two, or as many as three, may be con- 

 cerned in the formation of a single root, but, where more 

 than one is so employed, the intervening patches of phloem 

 are always very narrow. 



When a root is about to arise, those cells of the ray 

 which are situated in the phloem region of the bundle 

 undergo a remarkable series of changes, their nuclei 

 enlarge, they become richer in protoplasm, and at once 

 commence to divide irregularly in all three directions, 

 thus forming a small protuberance with a conical base, 

 which tapers gradually inwards towards the xylem, at the 

 margin of which it is continuous with the internal xylar 

 portion of the ray, and is here of necessity reduced to not 

 more than two cells in tangential width, and in most cases 

 to only one ; while the outer rounded apex of the papilla 

 is, on the other hand, composed at this stage of a convex 

 surface, with a perimeter of five or six cells, and generally 

 lies in contact with the inner surface of the pericycle, 

 which is easily recognisable by its fibres. At about this 

 stage, the thin-walled cells of the pericycle, opposite the 

 apex of the papilla, also undergo segmentation, usually 

 dividing, in the first instance, by radial walls, but so 

 irregularly as to render it a matter of great difficulty 

 to determine what part they take in the formation of 

 the root. 



Lemaire (7) mentions the occurrence of a somewhat 

 similar irregularity in the initial divisions, which give 

 rise to the adventitious roots of Tecoma radicans and Fic7ts 

 repcns, both of which further resemble Solanum, in that 

 their roots arise on the stems in a subaerial position, 

 though, in both cases, all the tissues of the root are formed 

 from the pericycle. 



In this connection the behaviour of the pericyclic fibres 

 is of some interest. Though normally the root in its 

 outward course altogether avoids these elements, it not 

 uufrequently happens that a single fibre or small group of 



