5'52 SECONDARF CHANGES. 



the other surfaces of the wall being flat and in close connection with one another (e. g. 

 Ginkgo, Sambucus and Lonicera) ; or the walls are rounded off on the greater part 

 of their surface, and the cells are therefore onlv in slight connection, forming, 

 especially when dry, a loose, powdery mass, e.g. Prunus avium, Pyrus malus, Robinia, 

 Betula. ^sculus, and Gleditschia. In the latter case the mass of complementary cells 

 is held together, owing to the fact that some layers of flat cells, firmly but not uninter- 

 ruptedly connected among themselves and with the adjoining loose ones, are always 

 formed alternately with some loose layers of complementary cells. 



In the complementary mass of Quercus Suber, which is also loose and powdery, 

 I did not find this arrangement ; the cohesion is here due to the fact that the whole 

 lenticel is enclosed in the tough, firm mass of cork, and thus protected from falling 

 to pieces, as will appear still more clearly below. 



The firmer layers in the loose lenticels appear independently of the limits 

 between the periods of vegetation; e.g. Fig. 221 shows two of them in a lenticel 

 taken from a this year's shoot of the Birch on the 5th of June. In older lenticels, 

 however, even in those filled with dense tissue, the formation of an uninterrupted 

 layer of cork over their whole surface occurs in many trees at the close of each 

 period of growth ; at the beginning of the next period complementary cells are then 

 again added by the meristematic layer. Where such a layer of cork appears at the 

 close of each period of vegetation (e. g. Ginkgo) it marks the limits of the annual 

 zones of growth. 



The uninterrupted layers of cork serve to shut off the internal intercellular air- 

 spaces ; the former therefore constitute the ' closing layers.' The closure is, however, 

 temporary, as the successive layers of cork are again burst by the subsequent formation 

 of complementary cells. 



In trees which form an autumnal closing-layer, the latter is, according to Stahl, 

 already present before the time of the fall of the leaf. The renewed formation of comple- 

 mentary cells below- it begins with the next period of growth, but need not immediately 

 result in rupture and the opening of aerial communication, as this obviously depends 

 on the relation existing between the pressure exercised by the new formation and 

 the resistance offered by the closing layer. As a matter of fact, judging from Haber- 

 landt's experiments \ this usually begins after the trees have completely expanded 

 their leaves, or even after the close of the flowering season in those in which the 

 flowers appear later than the leaves, although, according to the author's own judgment, 

 no certain conclusion can be drawn from these experiments alone, but more accurate 

 results must be sought by means of anatomical investigation. 



The outermost layers of the lenticel, for the time being, undergo, in proportion 

 to the progress of growth in thickness and phellogenetic new production, the same 

 passive changes as the layers of cork, namely desiccation and gradual decay. 



The productiveness of the phellogenetic layer in a lenticel is, as a rule, especially 

 in the centrifugal direction (centripetal with reference to the succession of the dividing 

 walls), greater than outside the lenticels; the latter therefore project as convex 

 bodies above the surrounding peridermal surface. Only in those trees, which like 

 Ulmus, Liquidambar, Euonymus europseus, and Acer campestre, form w'ing-like, 



» /.r. p. 16. 



