34 WRIGHT : THE GENUS DIOSPYROS 



of double rays. In tracing one ray from the younger to th<- 

 older wood it was first seen to be three cells broad, and very 

 gradually one marginal row was replaced by a row of fibres, 

 and finally, the double ray passed into a single ray, each 

 component having a tangential diameter equal to the sum of 

 those forming the double ray. It has also been noticed in 

 specimens of I), attenuata, that whenever the ray is more 

 than one cell broad the component cells are invariably of 

 the horizontal type, and at each end of such a ray the 

 components are of the vertical type with large tangential 

 diameters. 



It is to be noted that whenever the ray is more than one 

 cell broad the components have always a small tangential and 

 large radial diameter, and many sections show a double 

 ray superposed on a broad single ray, each ray extending 

 vertically through a distance of many cells. 



The medullary rays form large vertical strands of tissue 

 from twelve to over thirty cells in height, the components 

 having very variable dimensions and forming part of single, 

 double, and triple rays. 



The medullary rays, since they pass through all the 

 secondary xylem, come into direct communication with every 

 other type of cells, but the most striking communication is 

 with the wood parenchyma, which they so closely resemble 

 (particularly the vertical type) in structure, thin walls and 

 crowded contents. 



Form and size can be treated better in a combined descrip- 

 tion of the differentiation of medullary ray tissue. 



In nearly every species there is a well marked differentia- 

 tion into (a) horizontal cells characterized by large radial 

 and short tangential and vertical diameters, and (6) vertical 

 cells having short radial and large tangential and vertical 

 diameters. In many cases one feels strongly inclined to 

 introduce yet another type having approximately the form 

 of a cube, 



