i 9 o8] THIESSEN—DIOON EDULE 367 



donary strand leaves the vascular cylinder it is endarch (fig. 10). 

 As it passes upward and outward, the protoxylem elements recede 

 from the endarch position and are buried more and more in metaxy- 

 lem, that is, centripetal wood has appeared. When well up in the 

 tubular portion of the petioles of the cotyledons, the protoxylem is 

 surrounded equally on all sides by metaxylem, the xylem being 

 typically mesarch (figs. 11-13). From this point on the protoxylem 

 approaches more and more an exarch position, but the xylem does not 

 become completely exarch. Before the end of the bundle has been 

 reached it has become quite concentric, and it becomes very difficult 

 to determine which element was the first to appear (fig. 15) . Where 

 the bundle leaves the central cylinder the total xylem elements so 

 far as developed are at the innermost part of the bundle, and as the 

 bundle is traced upward they'recede from that position and occupy 

 one farther inward, until in the upper extremity of the cotyledon 

 thev occupy a position central to the whole bundle. Thus the trans- 

 position of the total xylem holds the same relation to the procambium 

 as the protoxylem holds to the metaxylem (fig. 9, a). 



The leaf traces also when leaving the vascular cylinder are endarch 

 (Jig. 16), and in passing outward and upward the protoxylem ele- 

 ments recede from the inner edge and are buried deeper and deeper 

 in the metaxylem; and well up in the leaf bases the xylem has become 

 mesarch (figs. 18, 19). Afterward the protoxylem approaches more 

 and more an exarch position, until at the transition between petiole 

 and leaf base the xylem has become completely exarch, with the pro- 

 toxylem lying immediately against the procambium (figs. 20, 21). 

 There is at this stage no centrifugal wood above this point; and there 

 is no secondary wood anywhere. The transition may be said to occur, 

 therefore, between the central cylinder and the leaf base, from which 

 point upward the strands are all exarch. It must be taken into 

 consideration that none of the leaves are as yet fully developed, only 

 the first leaf showing plainly the different regions (fig. 3). Fig. 9, a 

 represents the situation diagrammatically. 



It will be observed that only a very small part of the procambium 

 has been developed into xylem tissue in the whole length of the petioles 

 (figs. 16, 17). Cross-sections of foliar strands at a low region show, 

 besides the protoxylem and centrifugal metaxylem elements, a pro- 



