of a Urcdine, Pnccinia dispersa (Eriks.). 463 



nearly approaching the normal than, perhaps, the tubes afford, to see if 

 the influence of mineral supplies, or lack of such, at the roots affects 

 the questions before us. 



Seeds of four species of Brome were sown in pots of coco-fibre 

 (coir) thoroughly steamed in a steriliser. The seedlings all except 

 B. asper came up satisfactorily, but did not form strong plants 

 eventually. As the following Table VII shows, infection failed in all 

 cases, in spite of the fact that the seedlings were given a double dose 

 of spores. It is difficult to explain the negative results with B. molli*. 

 The temptation is to conclude that the host-plant, being starved of 

 mineral supplies, gave the infecting germ-tube no chance to establish 

 itself, but I think much more work must be done before this is 

 established. 



The following experiments, however, yielded results which seem to 

 show that the nature or amount of the mineral supply may after all 

 be not without effect. 



Small pots were packed with washed coco-nut fibre and sterilised 

 by steam, and " seeds " of B. niollis sown in twelve of them. Ger- 

 mination was started at 2? c C. in the artificial greenhouse and the 

 pots then placed in a well-lighted, ordinary greenhouse and watered 

 as follows : Two with Detmer's normal solution, and two each with 

 the same minus K, Ca, Mg, and PoO^ respectively, while two were 

 kept as controls watered with distilled water. 



On the 5th day the spears were up and the first leaf unfolding, 

 and one pot of each pair mentioned infected with spores from B. innlli*. 

 The parallel set were similarly infected, with spores from B. inolli*, 

 two days later, a different crop of spores being used. 



After four days' incubation under wet beakers in the case of the 

 first six pots and two days in the case of the second six all were 

 put out side by side in a sunny place under a large bell-jar. Two days 

 later the bell-jar was removed. The pots were placed well apart, 

 each in its own saucer, and equal carefully measured doses of the 

 minerals given at intervals. 



The seedlings grew admirably at first, but began to show the 

 effects of the mineral deficiency in a few days. Meanwhile the incuba- 

 tion period was being completed. The results appear in Table VIII. 

 There were, however, very marked points of difference l>etween the 

 appearances in this series 716 and those seen in perfectly normal 

 plants e.g., series 735 736 and 738. 



In the first place, none of the pustules or disease-flecks were very 

 large or well developed. They were, curiously enough, test on the 

 seedlings provided with water only. Next best on those with normal 

 mineral solution, and on those with the same devoid of Ca, and 

 (another curious fact) in one case on the plants deprived of Mg. On 

 those deprived of P_>0 5 or of K the development was distinctly worse. 



