358 BOTANICAL GAZETTE [November 



finally entering the leaf base. There is such a girdle for every leaf, 

 and every girdle must cut, on account of its course and location, every 

 other like girdle in two places. Every girdle receives on its inner 

 edge branches from the central vascular bundles which leave the 

 vascular cylinder at various places, and sends out branches from its 

 outer edge to other girdles. From this vague conception, which does 

 not at all agree with the drawings of Mettenius, the current text- 

 book accounts have been drawn. These accounts, however, do not 

 really interpret Mettenius, but are as far from his interpretation as 

 that was from the true situation. 



In his description of the structure of the bundles Mettenius was 

 more fortunate, and made a very important contribution. He says 

 that those bundles which are to leave the vascular cylinder are marked 

 off by broad medullary rays and are more definitely bounded than 

 those which continue in the cylinder; that on the inner edge they are 

 provided with spiral vessels (protoxylem), while the others are pro- 

 vided with reticulated cells in the same relative position. At the region 

 of the outward bend of the trace the vascular elements are grouped 

 with the spiral elements on the inner or upper side, immediately 

 bordering the reticulated elements beneath; and this structure the 

 girdle retains in encompassing the stem. Before entering the leaf, 

 however, a change in structure begins to occur, and is completed in 

 the lower part of the petiole; after which the bundle remains un- 

 changed up to the pinnae. The first indication of this change is the 

 appearance of thin-walled cells in the vicinity of the spiral (protoxy- 

 lem) elements, separating them from the reticulate elements. During 

 the further course of the bundle the spiral (protoxylem) elements 

 gradually move farther within, and the wood is now divided into two 

 parts by the thin-walled cells; the inner part developing its elementary 

 constituents centripetally, and the outer part centrifugally. Finally 

 the spiral (protoxylem) elements are found in the outer part of the 

 bundle, and the centrifugal part is still more reduced, while the 

 centripetal part has reached its maximum development. This struc- 

 ture of the bundle is retained in its further course in the petiole until 

 in the pinnae, where in Dioon edule the centrifugal wood is lost 

 altogether. Such is a very brief statement of the description of the 

 transition from centrifugal to centripetal xylem in the leaf traces, as 



