148 Prof. H. Marshall Ward. Effect of Mineral [Nov. 4, 



chemically purified by acid and washing, and stronger doses of the 

 mineral solution were employed. Each beaker received an equal 

 weight of the sand (1500 gr.), and of soluton (300 c.c), and 8 grains 

 of a species of Bromus, received under the name of B. pendulinus, 

 were sown in each. This species — which differs little, if at all, from 

 B. patulus — has been thoroughly investigated with regard to its 

 predisposition to attack by the uredo of Puccinia dispersa, and was 

 chosen because it has proved to be the most sensitive to infection of 

 all the forms I have in culture. The grains were hand-picked and 

 clean, but were not specially sterilised in any way. 



The beakers were treated as before, and on July 15 seedlings were 

 showing in all but one of the beakers, and by the 18th each of 

 thirteen had from seven to eight seedlings, two had four each, and one 

 none. By July 30 every beaker had from six to eight excellent seed- 

 lings, except one, and this remained barren to the end, possibly owing 

 to drowning of the grains, or too deep sowing. 



By this date there were marked differences in stature, and some in 

 colour, between the various pairs of sets ; the smaller seedlings in the 

 beakers deprived of all added salts, or deficient in nitrogen or in 

 phosphorus being particularly noticeable. Not only were the leaves 

 of these fewer and shorter but also narrower, and the whole plantlet 

 looked feebler in each case than corresponding specimens in other 

 beakers. Most of the latter had three or four leaves fully expanded, 

 with the fourth or fifth just peeping through, but these feebler seed- 

 lings only showed two mature leaves and a third one just beginning 

 to unfold, or three leaves with as yet no signs of the fourth. 



On July 30 I infected, chiefly on the second leaf,* with Uredo- 

 spores of the fungus grown on Bromus patulus, a form known from 

 experiments to infect this species of seedling very readily. The in- 

 fected plants were then placed under moist bell-jars, lifted in due 

 course, and treated as before. 



On August 8 several of the plants showed pustules, indicating 

 successful infection, and by August 12 it was possible to estimate the 

 degree of attack in each case, and to compare the relative virulence or 

 vigour of the pustules and spores developed. 



The essential facts are summarised in Table III. 



Here, again, certain concordant facts come prominently to light. 

 As before the infection is generally successful on just those plants, and 

 at just those spots where the inoculation occurred. In the apparent 

 exceptions (e.g., in g and i, Table III), I have no reason to doubt that 

 the extra-infection was due to the almost unavoidable splattering of 

 spores, or the contact of inoculated leaves with others, a particularly 

 easily incurred danger with this extremely susceptible species. 



* In several cases the first and third leaves also. 



