I I 2 
Ernest S. Salmon. 
signs of infection resulted when conidia were sown on species of 
Alchemilla, Fragaria, Spiraea, Agrimotiia, and Potcrium, on which 
the morphological species 5. Huviuli occurs. 
Hitherto no species of the genus Spliaerotheca has been 
investigated with respect to its adaptive parasitism. The above 
experiments prove clearly that in Spliaerotheca —as in the genus 
Erysiphe —specialization of parasitism has resulted in the evolution 
of “ biologic forms.” 
The case of the present species is of special interest because 
S. Humuli is the well-known Hop-mildew, and the cause of a 
disease which—especially in England—annually occasions con¬ 
siderable loss to hop-growers. If it can be shown that the form of 
the fungus on the Hop, is like that on Potcntilla, confined to its one 
host-species, or to species of one genus, the means of extirpating 
the disease must be sought in destroying all affected wild and 
cultivated hops, rather than in attempting to remove indiscrimi¬ 
nately all the mildewed weeds of a hop-garden. 
The Oidiurn of S. Humuli var. fuliginea (Schlecht.) on 
Taraxacum officinale and Plantago lanceolata was also used, and 
proved in each case to be a specialized “ biologic form.” 
The Oidiurn on Taraxacum officinale (see Table 6) produced 
virulent infection at once when sown on this host-plant, but failed 
to cause any infection on Plantago media, P. lanceolata, and the 
cultivated strawberry ( Fragaria sp.). 
A fact of considerable interest appeared in the two experiments 
in which the potted plants of Taraxacum officinale were inoculated. 
In the first experiment, on the twenty-first day (September 29) 
after inoculation, the fungus had produced on the six inoculated 
leaves many hundreds of perithecia among the still powdery Oidium- 
patches. By the twenty-sixth day (October 4) a great number 
of ripe dark-brown perithecia—each containing an ascus with 
ascospores—had been developed on the six leaves. On the twenty- 
ninth day (October 7) a quite young, scarcely fully grown leaf of 
the controls was observed bearing a small 0/<f/«/»-patch with a 
cluster of dark-brown perithecia in the centre. The abundant 
formation of perithecia on vigorous, quite young leaves, and at 
places which had previously borne a rich crop of conidia, show that 
for the production of perithecia other factors must be operative 
than those assumed by Neger. 1 In the second experiment a for- 
1 Flora, bXXXVI 11 ., p. 343 (1901). See also the writer’s remarks 
on the present subject in Bull. Torrey Rot. Club, XXIX., 
p. 19 (1902). 
