O. N. Purvis 351 



affected. The mechanical strength of the cell was, in consequence, 

 reduced during the early stages, but later no difference was apparent. 

 It is possible that the dryness of the season may account for the absence 

 of nitrogen effect in the later stages. During May when the rainfall 

 was greater, a weakening on the nitrogen plots was observed. 



B. In xylem. 



Plot 9. Full minerals + ammonium sulphate. 

 Date May 3Ist June 8th June 15th June 26th 



In the material collected from Plot 9 on May 16th no measurements 

 of the xylem were made as the vessels were crushed and distorted, and 

 accurate observation was impossible. 



The walls of the xylem appeared to be slightly thicker where nitrogen 

 had been applied, except on May 31st (Fig. 15). 



The lumen, however, was considerably greater where nitrogen had 

 been added than with minerals only, except on June 1.5th. This difference 

 was much more marked at the beginning and end of the period than on 

 intermediate dates (Fig. 16). 



The ratio of lumen to wall as seen in Fig. 17 indicates that where 

 nitrogenous manure was added the mechanical strength of the vessels 

 themselves was greater at the beginning and end of the period, but on 

 the intermediate dates, June 8th and June 15th, was rather less than 

 where minerals only were applied. 



Comparison of Plot 9 with Plot 10. 



A consideration of the effect of potassium salts in the presence of 

 added nitrogen is now possible. Plot 10, like Plot 9, receives a dressing 

 of ammonium sulphate and minerals, but here potassium salts are 

 omitted. 



From Fig. 18 it is clear that in the plots without added nitrogenous 

 fertiliser the walls were thicker in Plot 10, where potash wa^ withheld 

 than in Plot 9 where it was supplied, except on one occasion (June 26th) 

 when they were appreciably thinner, and on May 31st, when the means 



