N. E. Stevens — New Jersey Palmoxylon. 431 



large cells from near the origin of a root, in longitudinal sec- 

 tion. It will be noted that these cells are from three to five 

 times as long as they are broad and that the large pits are con- 

 siderably elongated. It is difficult to place these cells in any 

 of the recognized categories of lignified elements. They do 

 not seem to correspond exactly to either the "Kentia"or 

 " Kaphia" types of fibers described by Drabble (p. 435), but 

 may perhaps be designated as "stone" cells. 



The internal zone of the inner cortex usually consists of 

 three layers of rather thin-walled cells closely packed together 

 without intercellular spaces. These cells vary considerably in 

 size but are usually from one to three times as long as they are 

 broad. Compare fig. 13 with fig. 16. 



The endodermis is almost uniformly one cell thick, the cells 

 two to four times as long as broad; and even in the fossil the 

 radial walls appear markedly thicker than the tangential walls. 

 Compare fig. 13 with fig. 16. 



All the regions described for the fully developed root can be 

 made out in the smaller root shown in fig. 22. In the smaller 

 root, however, the various parts of the inner cortex are not so 

 clearly differentiated and all the cells have much thinner walls. 



Nuclei in cells of inner cortex. — Three longitudinal sections 

 showed the parenchyma cells of the inner cortex in a very 

 remarkable state of preservation. Indeed, the majority of the 

 cells of this region contained structure so characteristic in 

 appearance and so constant in occurrence that, if seen in fixed 

 material from living plant tissue, they would unhesitatingly be 

 described as the well-stained nuclei. 



Drawings of such structures would naturally afford no cer- 

 tainty as to their nature ; and the photomicrographs, figs. 18 and 

 19, are accordingly offered for what they may be worth as 

 proof. While not fully convincing in themselves these figures 

 are not wholly valueless. Practically every parenchyma cell 

 in the field showed a nucleus in some focus ; and in the figures 

 nuclei appear in the cells marked N as well as in some others. 

 It will also be noted that the triangular intercellular spaces 

 spoken of above clearly appear in these photomicrographs ; and 

 the probability of the cells having been " fixed " when in an 

 actively growing condition is further denoted by the occurrence 

 of pairs of cells which have apparently just been separated by 

 a cross wall {y, fig. 19). In these " daughter cells" the nuclei 

 are still close to the dividing wall. 



The Stele. — The pericycle is very readily distinguished as a 

 single layer, or sometimes locally as two layers, of rather regu- 

 lar cells inside the endodermis. These cells tend to be some- 



