68 CROWN RUST 
each of these forms would not infect the other hosts, so far as 
they were tried. It is doubtful if the two last forms are dis- 
tinct, as they were not tested sufficiently. 
A more complicated case is seen in P. coronata Corda, the 
Crown Rust. Not only is this divided by Eriksson and Klebahn 
into two sub-species, P. coronata and P. Lolii (=coronsfera), 
but each of these is still further sub-divided by them into 
biological ‘races. 
In P. coronata (ecidium on Rhamnus Frangula) they are: 
f. Calamagrostidis—on C. lanceolata etc. (also on Phalaris) 
f. Phalaridis—on P. arundinacea (also on Calamagrostis) 
f, Agrostidis—on A. vulgaris, A. alba, A. stoloniferc 
f. Holct—on H. lanatus, H. mollis 
f. Agropyri—on A. repens. 
She oe Poo 
The state of P. coronata which occurs on Dactylis has not 
yet been assigned to any of these forms. The last two of them 
seem doubtful and may not belong to P. coronata, and the first 
two of them may be identical, since each seems able to extend 
at times to the other host. 
In P. Loliz, the Crown Rust of Oat (ecidium on RA. cathar- 
ticus), they are: 
. Avenae—on A. sativa, A. fatua 
. Lolii—on L. perenne (perhaps also on Festuca elatior) 
. Festucae—on F. elatior, F. gigantea 
. Holci—on HA. lunatus, H. mollis 
. Alopecuri—on A. pratensis etc. 
. Glyceriae—on G. aguatica. 
To which of these the form of P. Lolii on Arrhena- 
therum should be assigned seems to be undecided. There are 
also Crown Rusts, of which little is known, on Melica nutans 
and Sesleria coerulea. 
Later (Arkw for Botanik, 1908, vol. viii) Eriksson revises 
his previous conclusions in regard to P. Lolti as follows, making 
eight races : 
Oo Pp wee 
re rh rh bh ph Ph 
1. ff Avenae—on Avena 
2. £. Alopecuri—on Al. pratensis, etc. (sometimes on Av. sativa) 
3. f. Festucae—on F. elatior, F. gigantea 
4. 
f Lolii—on Lolium perenne and other species (also on F. elatior) 
