22 A Monograph of the Erysiphaceae 



r 



host-plants of very different affinities. I have observed PJiyllactinia 

 corylea on hosts belonging to 48 genera, in 27 different families, 

 and Ery&iphe polygoni on 190 different species of host-plants, be- 

 longing to 89 genera. The latter species has been reported on 

 146 additional host-plants (some of which, however, must be 



considered doubtful). 



The members of the Erysiphaceae fall Into the following divi- 

 sions; 15 species and 2 varieties are confined to a single species of 

 host-plant; 15 species and 5 varieties to species belon'gingt to one 

 genus ; 4 species and I variety to those belonging to one family, 

 and 15 species and 3 varieties show little or no preference in select- 

 ing^ their hosts. In the species of the first division it seems reason- 



able to suppose that the connection between host and parasite 

 must be a very intimate one. In those of the last division it seems 

 probable that the connection is slight. 



The number of cases in which more than one species or va- 

 riety of mildew has been observed on one species of host-plant is 

 rather large ; on one host-species five different mildews have been 

 recorded in one case, four in four cases, three in 24 cases, and two in 

 163 cases. These numbers have been obtained by taking into ac- 

 count all the published records, and are therefore (especially with 

 regard to those cases in which two species oi mildew have been re- . 

 corded on the same host) probably far too high. In the present con- 

 fusion that exists concerning the hosts of many species it is impossi- 

 ble to give exact numbers ; probably, in the last case mentioned, 100 

 in the place of 163 would be nearer the mark. In many cases it is 

 probable that an error in identification has led to different species of 



ysiph 



For instance, 



iph 



Magnus, and E, p 



E. cichoracearum grows commonly on Achillea Ptannica, and it 

 may perhaps be doubted if the two other species— ii. taurica and 

 E. polygoni^ recorded on this host, really occur. It is needless to 

 give further examples, as these can be found in the host-index. 

 On the one hand it is quite certain that in a large number of cases 

 more than one species of mildew grows on one host-species ; on 

 ■ the other hand, it is equally certain that \r\ many of the recorded 

 cases of this kind wrong determinations have been made. It is hoped 



