430 JOURNAL, OF FORESTRY 



ble row. In all cases where the ducts are close enough together it is 

 usual for them to anastomose into a tangential network which not in- 

 frequently excludes everything but the rays, or portions of them, which 

 remain like islands in the tangential section ; in Daniellia there are, ac- 

 cording to Guignard (i), occasional radial anastomoses independent of 

 the rays. 



The cells surrounding these vertical canals are mostly parenchyma- 

 tous, at least at first, though subsequent lysigenous enlargement in some 

 cases may effect a change. The limiting or epithelial cells may be 

 thick or thin-walled throughout or thin only in the part facing the cav- 

 ity. In the conifers there is always a distinct epithelium, and the cells 

 of it are thin- walled in Pinus, but thick- walled in the others, with occa- 

 sional exceptions. There is no uniformity in the origin and develop- 

 ment of dicotyledonous canals, being schizogenous, lysigenous, or 

 schizo-lysigenous, depending upon the species. In the case of Copa- 

 ifera, Rhein found, according to Solereder (2, pp. 285-6), that "side by 

 side with the secretory canals, produced, so to speak, from the cam- 

 bium, others arise subsequently in the mature tissue of the wood, by the 

 schizogenous origin of a small intercellular space, which subsequently 

 becomes enlarged in a lysigenous manner." In the coniferous woods 

 the resin canals are exclusively schizogenous, and originate only in the 

 cambium. In some instances in dicotyledons (namely, Dipterocar- 

 pacese) the ducts may become filled with tyloses, corresponding to the 

 proliferation of the thin-walled epithelial cells into the ducts of Pinus. 



Radial ducts are characteristic of many species of dicotyledonous 

 woods, and vary from extremely minute and only discernible with a 

 compound microscope, to large enough to be seen with the unaided eye. 

 Their location is often disclosed by the discoloration of the wood from 

 their exuding contents, appearing as specks on the tangential section 

 and as lines on the transverse and radial. They are very numerous 

 and show most conspicuously in a wood from British Guiana known as 

 "duka" (Tapirira sp.), and are equally numerous, though smaller and 

 less conspicuous, in Schinus wienmannifolius, from Argentina. With 

 a hand lens, and sometimes without it, they are visible in Astronium 

 and Didymopanax. In quebracho (Schinopsis) they are fairly numer- 

 ous, but owing to the deep color of the wood are usually obscure without 

 a compound microscope. The writer has found the presence of radial 

 canals, as also the vertical ones, very helpful in the identification of 

 tropical woods. Owing to their irregular distribution, they may be 

 present in one section of a wood and not in another from the same 

 specimen. 



