Iiiicction. 31 



appear 011 tlie lower nodes of the axis were also infected as soon as tlieii' tips 

 were exposed, and tliey, too, soon showed swellings which developed into 

 smut pustules. Thus, wherever the tissues were young and tender, the ger- 

 minating conidia were able to penetrate and produce infection which, how- 

 ever, was strictly confined to the parts directly attacked. 



■>. Flower Infection. 



Infection throuj:;h the liower has only been recently experinientalh^ proved, 

 and on account of its practical importance, will recjuire to be fully considered. 

 For this mode of infection, those smuts are best adapted which have powdery 

 spores, and are, consequently, easily blown about by the wind. What are 

 known as Loose or Flying Smuts, fulfil these conditions perfectly, as the very 

 name indicates that they are readily scattered, and those of wheat and barley 

 are characteristic examples. 



As far back as 189(5, Mr. Frank Maddox, then Agricultural Experimental- 

 ist to the Council of Agriculture of Tasmania, had practically demonstrated 

 infection by the flower, and his account of it in the Agricultural Gazette is well 

 worthy of being reproduced here. He writes : — " I will now give the conclusions 

 I have arrived at with smut (of wheat) from the results of my experiments. 

 I have never been able to cause infection and reproduce the disease with spores 

 on the grain or in the ground, which I can so easily do with bunt spores to 

 reproduce bunt. " The only way I have been able to infect grain and reproduce 

 smut (which seldom ever fails) is by putting the spores on the ovary of the 

 plant at flowering time, about the same time as the pollen-grains are being 

 shed. The grain will mat ; re without the slightest signs of being diseased. I 

 have hit the time so well now that I may say I never have a failure. I think 

 this accounts for when I did fail, viz., the ovary was not forward enough for 

 the spore to get its seed-bed, or possibly, sometimes the spores were not 

 matured enough. The comparison of bunt and smut spores finding their 

 seed-bed are the very opposite. It is really wonderful to me how the smut 

 spores do, as the ovary is well protected by the glumes or chaff, and there is 

 only a short period that infection seems to be able to take place. " There is 

 no doubt that here we have a practical demonstration of the fact that infec- 

 tion by the loose smut of wheat occurs during the flowering period, and that 

 this is the first record of it. 



Next year, in 1897, Nakagawa' infected the flowers of wheat with the 

 matured spores of loose smut. The infected seeds were sown, and next year 

 the plants were found to be smutted. Soon afterw^ards, similar results in 

 flower infection with the spores of Ustilago tritici and U. nuda were obtained 

 by Hori', and he concluded " that the spores of those smuts which mature 

 at the flowering time of the host, and may be scattered easily by the wind, 

 will be retained in the inner side of the seed and give rise to the smut I'.isease 

 during the next flowering time of the host-plant." 



In 1903, Brefeld" had also proved flower-infection in the wheat, and 

 Hecke^, in 1904, the same in th ? barley. In 1905, Brefeld published a series of 

 carefully conducted experiments. Spore material of the loose snuit of wheat 

 carefully selected and preserved through the winter, was dusted on to the 

 stigma of the wheat-flower, just as in the operation of crossing, and micro- 

 scopic examination showed that the spores germinated, and that the germ-tubes 

 passed down the style into the ovary, where a fine network of mycelial fila- 

 ments were formed. But the ])lants thus treated exhibited no signs of disease 

 and produced strong healthy grains in the ear. When the grains, however, 

 were sown next season, with all necessary precautions against outside infec- 

 tion, the resulting plants were so badly affected that the entire inflorescence 

 was destroyed. .\s the same result was always repeated the conclusion 



