352 THE AMERICAN NATURALIST. [Vol. XXXV I II. 



four well marked divisions : ( i ) round to oval or transversely 

 oval; (2) rays broad, the cells oval to round, chiefly round; (3) 

 chiefly oval, and (4) rays narrow, the cells oblong to oval, chiefly 

 oblong. The broadly oval and thin-walled cells of Sequoia sep- 

 arate it from associated genera. In Picea the genus may be 

 subdivided accordingly as the cells are ( i ) variable, round, oval 

 or oblong; (2) equal and uniform, oblong or oval.^ Capressus is 

 similarly separable into groups. But more specifically it is not 

 difficult to separate C. arizonica and C. goveniana by reason of 

 their broad rays and very conspicuously transversely oval cells, 

 from C. pisifera with its round or oval cells and C. thyoides with 

 its narrow, oblong, rarely oval cells. In the genus Pinus atten- 

 tion is at once directed to P. miirrayana by the conspicuously 

 round or transversely oval, very unequal and variable cells. 



The interspersal of the tracheids often imparts a character- 

 istic appearance to the tangential aspect of the ray, especially in 

 the genus Pinus, and more particularly among the hard pines. 

 In this group the tracheids present very variable forms and 

 sizes. In such types as P. glabra they are small, oval or round, 

 and wherever they occur they give rise to a marked local con- 

 traction. In P. palnstris and P. aibensis they are commonly 

 oblong and not infreq^uently they become several times higher 

 than broad. As they are almost invariably narrower than the 

 associated parenchyma cells, they cause a local contraction 

 which sometimes extends over considerable distances. In P. 

 palnstris the predominance of the tracheids is carried so far that 

 the rays are chiefly composed of them, and it then becomes ap- 

 propriate to apply the term interspersed to the few parenchyma 

 cells. In all of the more highly organized rays of the hard 

 pines the appearance of the structure is so complex and variable 

 that a proper diagnosis can be drawn only when we take 

 cognizance of the principal aspects presented, and these some- 

 times amount to as many as four in number. 



A consideration of the various structural features thus dis- 

 cussed in their relations to classification will show that no other 



width; uniform^o the cells of all rays which Ire Vetty instantly of one form ; 



