THE SKKDL1NG AXD Ynr.v; PLANT 63 



but other arnin<vnionts occur. In the parts where 

 fewer vessels occur it is not uncommon to find a K<- 

 of radial rows of about six to ten fibres end in a single 

 parenchyma cell, and thus are formed short, tangential 

 rows of wood-parenchyma cells, intercalated, as it were, 

 between the radial rows of other elements (fig. 12, p). 

 It often happens, moreover, that reticulated and pitted 

 vessels are closely surrounded by wood-parenchyma. 



The secondary medullary rays exist as single radial 

 rows of cells, agreeing in form, &c., with the cells 

 of the primary medullary rays. In contact with one 

 another or with wood parenchyma their walls have simple 

 pits, but they have bordered pits where they abut on 

 tracheids or vessels. In winter these cells are filled with 

 starch. On tangential sections (fig. 15) it is easy to 

 see how the vertical groups of cells have the same origin 

 as the groups of wood-parenchyma cells the difference 

 being that the cambial cells which are going to be trans- 

 formed by horizontal divisions, &c., into vertical rows of 

 ray parenchyma, undergo repeated tangential longitudi- 

 nal divisions, and so continued radial rows are formed. 

 The cells of these rays are often much shorter than 

 those of the wood-parenchyma, yet all gradations occur. 

 The mother-cells may be very long, evidently corre- 

 sponding to two, and they may also divide in the radial 

 longitudinal plane, and the ray become biseriate. 



These secondary rays start (on the transverse section) 

 from the first large vessels, or from younger ones, or 

 they may start from other points. The ray may some- 



