250 



are large, nearly isodiametric, and with thin walls ; below the 

 epidermis are two layers of thin-walled cells without chlorophyll, 

 then four or five layers ofparenchymatic chlorophyll-holding; cells, 



r 



and then below these again several layers of colorless cells." At 

 the corners of the stem the rows of chlorophyll-holding cells are 

 broken by a cluster of collcnchymatic colorless cells. In a later 

 stage of development the central cells of this cluster form the 

 string of bast fibers which are pushed out by the growing wing. 

 It is with the growth and development of these bast fibers that a 

 point of interest occurs. A section cut at the distance of % 

 centimeter from the growing tip (Plate CIX. fig. i.) shows this 

 little central cluster of cells, with thick walls and small lumen ; 

 now a section at three centimeters from the tip shows that several 

 important changes have occurred. The two cylinders of color- 

 less cells now begin to take on a collcnchymatic appearance, the 

 walls having thickened considerably. The cells of the epidermis 

 are in a state of rapid growth and division, new cells being added 

 to the outer cylinder whose walls rapidly thicken, as may be seen 

 by sections at the distance of five and seven centimeters from the 

 tip. The cells in the center of the group in the corners are seen 

 to have increased in diameter while at the same time their walls 

 are much thinner. 



Now if we folK.w this same set of cells, cutting at various dis- 

 tances below this, we find them with extremely thick walls, so 

 that only a small point appears as the lumen. Their diameter is 

 about the same as at the distance of three centimeters, but in 

 whatever way the additional thickness has been acquired the 

 original wallas seen at the latter distance is perfectly evident as a 

 dark line marking the outline of each cell. . 



Here, if we accept Strasburger's theory, that growth in sur- 

 face takes place by stretching, we have an apparently consistent 

 explanation of the processes which have taken place. The thick 

 walled cells in fig. i c, have changed to thin- walled wider 

 lumened cells as seen in fig. 2, by using up the material already 

 in their walls. By this process of stretching the diameter is in ■ 

 creased until the proper size for the bast fibre is obtained. Then 

 by a succeeding process this thin wall is added to until hardly a 

 trace of cavity remains. If, on the other hand, as in the case of 



