370 BROOKLYN BOTANIC GARDEN MEMOIRS 
3. Puccinia Centaureae (Mart.) Hasler. 
a. Scabiosae on Centaurea scabiosa. 
b. Nigrae on Centaurea nigra. 
c. Nervosae on Centaurea nervosa. 
d. Transalpinae on Centaurea transalpina, C. alba, C. austriaca, 
C. Jacea var. longifolia, C. nervosa, C. nigrescens and C. 
bhrygia. 
Puccinia chaerophylli Purt. Semadeni (139, 140) claims that a 
race of this rust occurs on Anthriscus silvestris and another on Chaero- 
phyllum aureum. 
Puccinia Epilobii-tetragoni (DC.) Winter. Dietel (29) was unable 
to infect Epilobium hirsutum with aecidiospores from E. tetragonum. 
This is the only indication of specialization in this rust. 
Puccinia Geranii-silvatici Karst. This rust has been reported on a 
few species of Geranium in widely separated localities. It occurs 
commonly on Geranium silvaticum in Europe. Its restriction to widely 
separated regions has led to the suggestion that the rust consists of 
geographically specialized races. Jacob (66, 67) has shown by cul- 
tural experiments that there is no specialization in Europe, for both 
Geranium silvaticum and G. rotundifolium are readily infected by 
teleutospores from the former. 
Puccinia Helianthi Schw. Arthur, Kellerman, and Jacky have 
made inoculation tests using teleutospores of this rust. Jacky (65), 
using teleutospores from Helianthus annuus, infected’H. annuus, H. 
cucumertfolius, and H. californicus, but failed to infect H. maximiliant, 
H. multiflorus, H. rigidus, H. scaberrimus and H. tuberosus. 
Kellerman (74, 75) obtained negative results with teleutospores 
from H. annuus on eighteen species of Helianthus. His results with 
teleutospores from H. ambiguus were negative on twelve species and 
also negative with teleutospores from H. decapetalus on eight species. 
He found, however, that teleutospores from H. mollis infected H. 
annuus and H. mollis but not fourteen other species; teleutospores 
from H. grosse-serratus infected H. annuus, H. decapetalus, H. giganteus, 
FI. grosse-serratus, H. Kellermanit, H. orygalus and H. tracheifolius, 
but not H. maximiliani nor H. mollis; teleutospores from H. tuberosus 
infected only H. annuus. 
Arthur (3, 4, 5, 6) secured the following results: (1) Teleutospores 
from Helianthus mollis infected H. annuus, H. hirsutus, H. mollis, 
H. occidentalis, H. strumosus and H. tomentosus, but not H. grosse- 
serratus, H. Kellermani, H. laetiflorus, H. longifolius, H. orygalus and 
HT, tuberosus; (2) teleutospores from H. grosse-serratus infected H. 
annuus, H. grosse-serratus, H. maximiliant and H. tomentosus but not 
FH. decapetalus, H. hirsutus, H. laetiflorus, H. mollis, H. occidentalis, 
