SECONDARY CHANGES OUTSIDE THE ZONE OF THICKENING. 565 



quence of its own growth, and of constant additions from the phellogen ; it first bulges 

 out the epidermis, and then breaks through it by a longitudinal crack passing through 

 the stoma, or through one of the numerous stomata, and then protrudes through the 

 gap. The complementary cells coming out through the crack, which is constantly 

 opening wider, then dry up, together with the shreds of epidermis adhering to them ; 

 they form those elements of the complementary mass which were mentioned above 

 as not being radially arranged, and as peculiar to stomatal lenticels. 



The beginning of the formation of lenticels takes place as a rule on the young 

 shoot at an early period, before longitudinal extension is complete, and before the 

 formation of other periderm. Indeed the latter as a rule starts from the edges of the 

 lenticels, as soon as the phellogenetic layer is formed in the latter, and is continued 

 thence over the surface of the shoot. Exceptions to this rule are rare, and apparently 

 peculiar to individuals ^. It is true that the two processes often follow one another 

 immediately, so that the appearance of periderm and lenticels may be spoken of as 

 simultaneous, without any great eiTor. In shoots with a long-lived epidermis on the 

 other hand (Sophora japonica, Rosa canina, Negundo, and Acer striatum), the lenti- 

 cels already appear in the first year, and thus long before the further extension of 

 the periderm. 



It is obvious that according to the epidermal or sub-epidermal origin of the 

 periderm, special differences must occur with reference to the conditions described, 

 especially as regards the connection of the segments of the phellogen with one another. 

 For these Stahl's work may be referred to. 



In the superficial periderms, a// oj-iginal kfilicels, so far as is known, are formed 

 below stomata in the manner described. 



Secondly, the formation of lenticels occurs independently of stomata, on periderms 

 either when first developing, or in older stages, owing, as it may be shortly expressed, 

 to the fact that the phellogenetic layer of meristem forms, on limited areas, lenticel- 

 tissue instead of ordinary periderm. If this begins after layers of cork are already 

 present, the latter are burst by the growing lenticel. A more minute description of 

 these processes is superfluous after what has been stated above ; for some special cases, 

 and also for the peculiar formation of lenticels on the sites of the insertion of the 

 leaves in Abies pectinata, a case which does not strictly belong to this category, 

 reference may be made once more to Stahl's work. 



Lenticels independent of stomata arise on the internal periderms, both on those 

 first formed and on the succeeding ones, simultaneously with the origin of the rest 

 of the peridermal layer (with the obvious exception of plants wholly destitute of 

 lenticels). 



According to Stahl's observation on Pyrus Malus, and Haberlandt's enumera- 

 tions to be mentioned below, they may also be successively formed anew, between 

 the pre-existing ones, on older periderms which have been growing for some time, 

 whether these be superficial, or of endogenous origin. 



The latter new formations increase the number of lenticels on those periderms 

 which follow the dilatation. Those appearing on successive periderms replace those 

 lenticels which are lost when the bark is separated. If the bark is thrown off in scales, 



' See Stahl, I.e. p. 23. 



