316 Professor Marshall Ward, On the Question of 
In the annexed table (Table III.) I have put together the 
results of 1816 infection- experiments with these pot seedlings, 
arranged in such form as to show at a glance their significance in 
the present connection. 
The results point without doubt to the conclusion that both 
the source of the spores employed and the specific peculiarities of 
the Brome inoculated are important factors in infection. For 
instance, uredospores derived from B. sterilis successfully infected 
B. sterilis in 68 cases out of 84 (81 °/ Q ), B. madritensis in 38 cases 
out of 61 (62'3 %), and B. maximus in two cases out of 82 
(2'4 °/ o ), whereas they failed entirely to infect B. erectus [60 J ], 
B. secalinus [77], B. velutinus [71], B. arvensis [72], B. mollis [84], 
B. racemosus [72], B. commutatus [72], B. interrupt us [72], B. 
Schraderi [9], and a number of others. 
Now B. madritensis is closely, and B. maximus distantly related 
to B. sterilis , and all three come into the group Steuohromus. 
The evidence goes to show that the spores produced under the 
influence of the species of host B. sterilis (Steuohromus group) 
have been so modified by the circumstances of their nutrition and 
rearing, &c. that they can successfully attack other host-species of 
the same (Steuohromus) group, but are unable to overcome the 
obstacles to infection presented by the Bromes of the Festucoides 
(e.g. B. erectus ), Serrafalcus (e.g. B. mollis, &c.), or Geratochloa 
(e.g. B. Schraderi) groups. 
Similarly with spores cultivated on B. mollis (Serrafalcus), 
these attacked B. mollis successfully in 60 cases out of 85 (70‘6 °/ 0 ) 
and other members of the same cycle of relationship (Serrafalcus 
group) in proportions diminishing more or less with the closeness 
of their relationship ; but only in one case out of 31 did a pustule 
arise on B. erectus (Festucoides), and only in four cases out of 90 
did infection succeed in B. sterilis (Steuohromus) as a consequence 
of sowing spores from B. mollis on these species. 
Here arises the question, Is this a case of spores raised on 
B. mollis adapting themselves to B. sterilis and B. erectus ; or of 
the latter proving individually less resistant than their species 
generally to the infection ? 
Even more striking were the results with spores which had 
been reared on B. secalinus (Serrafalcus). They infected B. seca- 
linus 16 times out of 16 trials (100 °/ 0 ) and B. macrostachys 
five times out of five trials, but were without result on B. inermis 
(Festucoides) [5], B. sterilis (Steuohromus) [18], B. Schraderi 
(Geratochloa) [9], and so on ; and here again the only exception 
to the generalisation that the spores from B. secalinus cannot 
infect any Brome out of its own group Serrafalcus was a single 
These numbers in square brackets denote the number of attempts made. 
