344 Tlic Effect of Potassium Salts on Dactylis glomerata 



just outside the bundle, but was separated from it at other points by 

 thin plates of assimilatory tissue. 



AVithin the sclerenchyniatous zone is a large-celled ground tissue, 

 through which the larger bundles run. The sharpness of the division 

 between sclerenchyma and cortex varies in the different specimens. 

 The centre of the stem is occupied by a large lysigenic cavity which 

 extends throughout the length of each internode. 



The bundles are collateral, and are surrounded by a badly defined 

 •sheath, whose cells are slightly lignified ; the xylem consists as a rule of 

 one or occasionally two protoxylem elements, two large metaxylem 

 vessels and a few smaller elements. As no significant difference was 

 observed in the general appearance of the sections obtained from 

 different plots, more detailed measurements of cells were made to find 

 if any finer difference existed. Attention was concentrated on the cells 

 of the sclerenchyniatous zone, whose function is almost entirely mechan- 

 ical, and of the large metaxylem elements, which though not to any 

 extent mechanical in function, have become lignified, and provide 

 material which might throw light on the problem in hand. Measure- 

 ments were made of the thickness of the cell-walls, and the diameter 

 of the lumina, while the ratio of lumen to wall was calculated for each 

 cell. 



In most cases ten cells or vessels from each slide were measured, but 

 occasionally where the material was insufficient, twenty measurements 

 were made of some of the sections. In either case, the total number of 

 readings was eighty for the xylem, and fifty for the sclerenchyma, and 

 these readings were taken from a number of individual plants varying 

 from four to eight. A few sections showed marked differences from others 

 from the same plot, and were rejected as anomalous. 



The readings thus taken showed considerable variation among 

 themselves; in one case, for instance, the width of the lumen in scler- 

 enchymatous cells ranged from 4/x — 16/x (Plot 7, June 26th), slide e. 

 This variation is unavoidable, but the validity of the result is consider- 

 ably enhanced by the large number of readings taken, and where the 

 "probable error" has been calculated, the significance of a difference 

 between two means can be decided: a significant difference is one that 

 is considerably greater than the sum of the "probable error" of the two 

 means. Thus, the means of the fifty readings for the diameter of the 

 lumina of sclerenchyma from stems of Plot 7 material (where full 

 minerals, including potash, were applied) are as follows: 



