Nos. 451-542.] AJVATOMV OF THE CONIFERALES. 



549 



canal is found to be more or less continuous, though it presents 

 frequent constrictions, and it is thereby reduced to very narrow 

 dimensions. It is this feature which causes the canal to exhibit 

 such marked variations in size when seen in transverse section. 

 The epithelium cells are narrowly cylindrical and rather long 

 and thick-walled, as well as somewhat strongly pitted. Out- 

 wardly they become much longer and relatively narrower, and 

 they eventually merge with the surrounding parenchyma tra- 

 cheids by which they may also be replaced. 



In Larix the same features of contiguity and coalescence 

 may be observed, except that in L. occidcntalis the resin pas- 

 sages sometimes form into . ,a\ 



cheids, and like the paren- 

 chyma tracheids from which 

 they may be separated with 



difficulty. The latter, therefore, which are absent from the 

 summer wood, can be distinguished from the elements of the 

 spring wood only when the pits on the terminal walls (Fig. 47 

 pr. t.) are brought into view, or, more rarely, when the pits on 

 the tangential walls are in evidence. Thyloses rarely occur, and 

 so far they have been noted only in L. occidentalism where they 



continuous zones of imper- 

 fectly organized structures 

 with the aspect presented in 

 Tsuga martcnsiana. The epi- 

 thelium is always well defined 

 (Fig. 47) and it consists of 



cells. The cells of the first 

 row are small, very variable 

 in form and size, thick-walled 

 and more or less strongly 

 flattened radially. They are 

 also commonly resinous and 

 more or less strongly pitted. 

 When there is a second row 

 of epithelium, the cells are 

 essentially like the wood tra- 



