NOTES ON THE ERYSIPHACEAE 203 
from “ Paraguay, May 1883, nr. 3815” sent by Professor Spegaz- 
zini as “P. suffulta (Reb.) Sacc.” [= P. corylea] is of interest from 
the fact that on the same perithecium we find, in about equal 
number, simple and branched penicillate cells (p/. ro, f. 14-15). 
[This is the plant recorded by Spegazzini, in “ Fungi Guaranatici ” 
(Anal. Soc. Cientif. Argentina, 20: 66. 1886) as growing on Colli- 
guaya Brasiliensis. This last name must therefore be expunged 
from the list of host-plants of P. corylea, as Professor Spegazzini 
has now written “ Zxcoecaria” on the label of the specimen. | 
It is, however, when we observe the variation which occurs in 
the shape of the penicillate cells in examples of P. corylea from 
different parts of the world that we obtain the most convincing 
evidence as to the worthlessness of the characters given as distin- 
guishing “P. clavariacformis.” 
Taking first examples on various hosts in the United States, 
We find that a perfect series of intermediates exists between forms 
having branched or unbranched penicillate cells. In the specimens 
in Rav. Fung. amer. exsicc. no. 624, on Alnus serrulata, we find 
“clavarioid” branching in comparatively short cells (f/. 70, 
I. 16-78), on Corylus Americana similarly shaped cells occur, on 
Alnus serrulata (from South Carolina) much branched cells ( f/. 
10, f. 16-18 ), quite similar tothose of “P. clavariaeformis,” are 
found intermixed with unbranched ones ; the same may be ob- 
Served in the specimens in Ellis’ North American Fungi, no. 1327, 
(2), on Corylus Americana (from Pennsylvania). Branched penicil- 
late cells, intermixed with unbranched ones, also occur in speci- 
mens on Carpinus Americana (from Illinois) (see p/. 20, f. 79), 
On Vaccinium stamineum (from West Virginia) and on Cornus 
Nuttallit and Quercus (pl. rr, f. x). The most remarkable North 
American form of P. corylea, however, occurs on Catalpa syringae- 
Solia, from Aiken, South Carolina (Roumeg. Fung. select. exsicc. 
n©. 4756). Here the branched stem-cell is, comparatively, ex- 
—) broad, measuring up to 35 across (f/. rz, f. 2-5). If, 
Owing Neger (see above) we took such a character as the length 
Of the stem-cell] of the penicillate cells as of specific value, then we 
should be compelled to make another “ species’ of a plant pos- 
: S€ssing such remarkably shaped stem-cells as these found on the 
J | Pbyllactinia on Catalpa. 
