8. P. Armstrong (51 



if possible, to carry the experiments out on a larger scale, and over a 

 longer period than had previously been done. 



(6) Rust scale adopted. 



In recording the severity of rust attack, an arbitrary scale was 

 employed similar to that used by Eriksson (4) and Biffen(2). In the 

 present case the relative extent of attack was indicated as follows: 

 = a rust-free plant; l=a slight attack; 2 = a moderate attack; 

 3 = a bad attack ; and 4 = a very severe attack. When examining the 

 F^ crop, it became apparent that a considerable proportion of the plants, 

 though not completely rust-free, showed only the merest traces of 

 attack. It was thought advisable to distinguish these from the dis- 

 tinctly rusted individuals grouped under grade 1, and they were therefore 

 given a special mark (Ix). 



Such a scale, of course, can only indicate the relative extent of rust 

 attack in a given season. In an exceptionally bad rust year (e.g. liHO), 

 plants placed in grade 3 may be as severely attacked as those which 

 are placed in grade 4 in another season; and similar seasonal fluctuations 

 may occur among the other grades. But, although the method is not 

 perfect, it appears to be the only practical one that can be used under 

 field conditions where thousands of plants have to be examined in a 

 comparatively short time. Further, this scale has been found sufficiently 

 reUable for the purpose in view, and it has therefore been adhered to 

 throughout the whole of these investigations. 



(c) The soil. / 



The soil in the cages (on which all the cultures were grown up to 

 the end of 1919) may be described as a medium gravelly loam. A sample 

 of the top soil to a depth of seven inches was taken in September 1919, 

 and the analysis showed that it contained an ample supply of plant 

 food materials in an available form. 



Apart from a light top-dressing of superphosphate apphed in the 

 spring of 1917, it had received no manure since 1910. Since the last- 

 mentioned year it had been hand-dug, and always carried a cereal crop. 



(d) Weather records. 



Records of the weather are given in Table I. These were taken at 

 the University Botanic Garden about a mile and a half distant. In 

 addition, daily notes were taken of the weather conditions at the Uni- 

 versity Farm. 



