24 BOTANICAL GAZETTE [july 



In Dianthera americana, the two chief types of abnormality 

 found in the family occur, that is, the astelic condition and the 

 medullary vascular tissue. The former is evidently not in any 

 way dependent on the latter, since in the seedling the plant is 

 frequently astelic when there is no medullary tissue developed. 



In the terms of the stelar theory, Dianthera americana in its 

 early stages is monostelic. In the young seedling the first endo- 

 dermis differentiated, surrounding the nodal ring, corresponds to 

 the inner layer of the cortex. The stem apex is rather broad and 

 flat, but the three histogen layers of Hanstein can usually be made 

 out. From the plerome is developed a central cylinder, all that is 

 within the endodermis. The parenchyma of this stele corresponds, 

 therefore, to pith and medullary rays. 



This seems clear enough in the early stages of Dianthera and 

 also in Justicia. In the latter this condition persists permanently; 

 there is always a well marked cortex and central cylinder, with its 

 pith. In the internode of Dianthera americana, on the other hand, 

 an endodermal sheath about each bundle is initiated, passing 

 around the bundle and inclosing on the inner face of the latter a 

 mass of parenchyma. From our previous interpretation, this 

 endodermis differentiates on the sides of the bundles from the 



hymatous cells of the medull 



from 



endodermis 



m 



Now let us examine the fate of the parenchyma 



derived from the plerome. That part 



within the endodermal 



becomes so mo 



as to appear exactly the same as the cortical tissue. At this 

 mature stage, then, there is no visible differentiation between the 

 cortex and the pith, except within the endodermis. And yet we 

 have seen that they had an entirely different origin, the one derived 

 from the plerome, the other from the periblem. Unless one leaves 

 out of account the different origin, and compares only the mature 

 structures, one is certainly not justified in saying in this case, as 

 Van Tieghem and Duliot (12, p. 275) say in their definition of 

 astely, that the bundles are "directement nlonses dans la masse 



