in the Brontes and their Brown Rust . 281 
B. mollis and B. sterilis gathered on June 1, from widely 
distant localities and kept in separate tin boxes till June 3, 
on the drops of water hanging from the water-pores at the 
tips of the first leaves. This method is quite successful in 
such cases, as direct and repeated observations have shown : 
the germ-tubes soon make their way to the stomata of the 
emerging leaf, and infection results near the apex only. 
Three pots of each species (or variety) were sown. One 
of each was infected with Uredospores from B. mollis, another 
with those from B. sterilis , while the third series was not 
infected, but kept as a control set. Each series was kept 
under a large bell-jar, with damp filter-paper in it, and left 
thus covered at a western window for two days. After that 
the whole of the series was put outside, in a sheltered spot, 
but still covered with the bell-jars. The garden used was 
well away from any plants of Bromus , and no diseased plants 
were allowed in it except such as I brought there. At the 
end of a week they were fully exposed to the weather. The 
latter had been warm and sunny during the first week, but 
by June 12-13 it turned cold and some rain fell, and several 
days of dull or wet weather, often very cold, followed, varying 
as to wind which was sometimes bleak and dry. The prin- 
cipal assumptions entertained here were: (1) that carefully 
selected clean ‘ seeds/ germinated in new pots under glass at 
a distance from any known source of infection, would give 
me host-plants free from contamination ; (2) that by gathering 
the spore-material from places some distance apart, and 
keeping it in each case quite separate in closed boxes, tubes, 
&c., it should be possible to ensure freedom from accidental 
infection ; (3) that by keeping the inoculated plants under 
glass until infection was well assured, it should be possible 
to avoid accidental infection ; because, even assuming that 
wind-blown spores reached the plants after subsequent ex- 
posure, the resulting pustules would appear so late that 
I could discriminate between the consequences of the primary 
and of the secondary infection ; (4) there was the further 
assumption that the artificially induced infection would occur 
