RECENT WORK, ETC. 475 



that even the dermatogen is not constantly distinct in 

 Gymnosperms, the outermost apical layer undergoing 

 periclinal divisions and thus contributing to cortex as well 

 as to epidermis. Neither does Koch find any distinct 

 initial group for the central cylinder ; on the contrary, he 

 finds that the supposed plerome gives rise to the pith only, 

 which is the first tissue to be differentiated, the procambial 

 strands arising from the so-called periblem. This is to 

 some extent a return to the views of Sanio and Russow. 



A more recent paper by Koch, already referred to 

 (15), is of great interest from this, as well as from other 

 points of view, and deals the heaviest blow at the theory 

 of Hanstein which it has yet received. So long as the 

 evidence against this theory was merely negative, depend- 

 ing on the frequent impossibility of distinguishing the 

 histogenetic layers at the apex, it was always open to its 

 supporters to argue, with Van Tieghem (5, pp. 697 and 

 776), that our conclusions must be drawn from the cases 

 where the layers are distinct, and that their want of dis- 

 tinctness in other cases is merely due to displacements of 

 the initial cells. 



The special merit of Koch's work consists in the fact 

 that he has compared similar growing points of the same 

 plant, at different stages of development of the shoot. He 

 finds that while, in Angiosperms, the dermatogen remains 

 distinct, the number of layers of " periblem " varies accord- 

 ing to the stage of growth. Thus in the lilac (15, p. 401), 

 if a growing point be examined between two periods of 

 leaf-formation, a single initial layer of " periblem " is found. 

 When, however, a new pair of leaves is in process of 

 origination, this layer undergoes periclinal divisions into 

 three or four layers, the apparent limit between periblem 

 and plerome becoming less distinct. At the next phase, 

 when the origination of new leaves has ceased for the 

 moment, we find the one initial layer again, and so on. 

 Similar results were obtained in Berberis (15, p. 412). 

 Now it is evident that this variation in the number of 

 apparent periblem-layers, proves that they are not perma- 

 nently distinct as histogenetic initials. The same layer 



