159 



turning brown (PI. III, fig. 16 a, at k). In this way arise 

 the brown streaks on the surface of the knobs which 

 flnally by extension in tangential direction of this suberizing 

 process becomes entirely brown. A spécial suberizing 

 meristem, a phellogen, is not formed. 



The regular structure of thèse cell-hills is lost as soon 

 as the differentiation of a meristem commences. Some cells, 

 assembled in a small group, then enter a new stage of 

 strong growth, which makes them conspicuous in the 

 préparations by a more rounded form amidst the adjoining 

 cubical cells (PL III, fig. 14, m.a). A number of the sur- 

 rounding cells are compressed by the pressure which thèse 

 primordial cells cause by their growth anddie. (fig. 14pr. c.) 



Soon the primordial cells divide into a number of small 

 filial cells with extremely thin walls and dense contents, 

 after which the primordium has become meristem (PI. III, 

 fig. 15 m.). 



For answering the question in what place in a knob 

 the meristem is formed and which is the descent of the 

 initial cells, we hâve the following data. An otherwise 

 415 to 450 fi thick leaf-tip had by local swelling, to about 

 840 /i, formed a knob, which by a small dépression in the 

 middle was, so to speak, divided into two halves, each of 

 which contained a primordium of a meristem. One of thèse 

 primordia is the one figured in fig. 14. The surface of 

 the knob was entirely suberized to a fairly considérable 

 depth. In one half the primordium lay 220 f* below the 

 top of the knob and its cells in ail probability descended 

 from the subpalissade cells, in the other half the primor- 

 dium lay 180 /M below the surface and was of the same 

 origin as in the former case. 



While in another case a primordium was noticed which 

 genetically belonged to the original spongy parenchyma I 

 found in a small knob which was still covered by an 

 intact epiderm and did not rise more than 85 /" above 



