
| P. glumarum as to be only distinguished with difficulty. It has its 
eecidiospores on Anchusa in autumn, those of P. glumarum being 
unknown. P. dispersa has four forms—on (1). Rye, (2) Wheat, 
(3) Twitch, and (4) on the Bromi. 
are very much smaller. ‘The teleutospores are so like those ‘of 
} Puccinia simplex (a species first described by Kornicke) has been / 
) | known for many years. It is confined to Barley, and is common in 
| England ;, it is easily recognised by the large number of its teleuto- 
spores, which have one cell abortive. 
Secondly, there are the crown rusts, P. coronata and coronifera 
whose life-history has been worked out by Klebahn. The former ha 
its zcidiospores upon Rhamnus frangula, and is met with in five 
forms, according to Eriksson, viz., 1, Calamagrostidis ; 2, Phalaridis; 
3, Agrostidis ; 4, Agropyri; 5, Holci. The latter (P. coronifera) ha 
its zecidiospores on R. catharticus, and occurs in 6 specialised forms, viz. 
1, Avene; 2, Alopecuri; 3, Festuce; 4, Lolli; 5, Glyceria ; 
6, Holci. 
Eriksson draws attention to the fact that the extension of rus 
(uredo) from plant to plant is after all usually confined to a few yards 
For instance, P. graminis on Twitch is one of the commonest forms — 
in Sweden, as it is in England ; if one searches in the vicinity of a — 
Barberry-bush, one is almost sure to meet with it, but the affected 
plants are not found many yards from the bush. ‘This form, it will — 
be remembercd, also occurs on Barley and on Rye, but although it is — 
common enough on Twitch it seldom spreads, and then only to 4 : 
limited extent to the above-named cereals. A further instance, show- 
ing how limited these uredospores often are, is given, which occurred — 
; at the experimental station at Stockholm in 1894, where a plot of — 
Ha Horsford’s Pearl, a variety of Wheat specially liable to P. glumarum, s 
il: was grown between five other varieties. On the Horsford the char- — 
acteristic uredo first showed itself on May 11th, and in forty-three 
days had attained its maximum, but at the end of that period only 
three of the surrounding plots showed any signs of this rust, and these 
only to a very slight degree. 

ee 
The germinating power, both of the uredo and of the ze cidiospores, 
is often but small and at best capricious. Eriksson considered, upon 
reflection, that since Roman times, hot days and cold nights have been 
considered favourable to mildew, he resolved to try the effect of placing 
these spores upon ice for some hours, and by doing this he succeeded 
in arousing them to germination. The distance to which the zcidio- 
spores are capable of spreading the fungus from an individual zcidial 

