204 THE BOTANICAL MAGAZINE. [Vo i. xxiv. No. 284. 



with the inner zone of transfusion tissue of the present pines, 

 in which its development is however not so much pronounced 

 as in Prepinus, or with the latter region together with the 

 region of sclerenchymatous fibres just inside the latter region 

 and at the outer limit of the phloem. The inner sheath of 

 Prepinus japonic us is probably comparable with the sclerenchy- 

 matous fibre zones in the two species of pines just described. 

 Whether we have here in the lignified tissue at the ventral side 

 of each xylem to do with any of the elements of a centripetal 

 wood or not, is not easy to be decided. The two or five rows 

 of radially arranged cells above mentioned in this region or 

 any of the group of cells between those rows may or may 

 not represent the centripetal elements of wood. There is one 

 point, however, to be noticed about these two or five rows of 

 cells. The fact is that they are direct extension of the 2 or 5 

 'medullary rays' or ray plates of each xylem portion, so 

 that the cell at the beginning of each radiating row of cells is 

 directly fitting the innermost ray cell, which is unlike the cells 

 of a medullary ray of the stem, much elongated vertically. 

 The cell walls of both are lignified ; the ray cells have 

 usually cell contents, while the cells of radial rows in question 

 have none. Thus these radial rows of cells are placed in 

 alternating rows with the protoxylems of the centrifugal xylem 

 which are sometimes destroyed as is usual with this part of 

 xylem and leave gaps in the tissue. 



It may be a question, whether any tracheal elements on the 

 ventral side of a centrifugal xylem developed regularly, in direct 

 continuation with a medullary ray of a stem or a medullary 

 plate of a leaf in the same plane with the latter, and not 

 starting from one of the protoxylem elements, can be taken for 

 elements of a centripetal xylem. Most cases of usual mesarch 

 bundles seem not to give any convincing data for this point, as 

 their elements are usually rather irregularly arranged. The 

 Sigillarian wood seems too not to be comparable, as we 

 have there to do with a combination of a primary and a 

 secondary wood. 



The presence of a small initial cell near the protoxylem of 



