CEYLON FUNGI. 
105 
the subject was reopened by Massee in Grevillea, XXI., p. 118 
(June, 1893). Massee describes Triphragmium davellosum B., 
quoting Gard. Chron., 1857, as the original record, and citing 
all the references given above, except that of Anderson. He 
does not refer to the type specimen, but cites “ Rab.-Wint. 
Fungi Eur., No. 2918,” a specimen on Aralia nudicaulis, 
collected in New Hampshire, United States. Massee states 
that Triphragmium Thwaitesii is identical with T. davellosum, 
and expresses doubt whether both are not the same as 
T, eehinatum Lév. ; he does not state which specimens of 
T . Thwaitesii and T. davellosum were compared, and hence 
there is an element of doubt in this determination, since the 
fact that Ceylon specimens of these two are identical does not 
assist in any way. Nor does he say anything about the host- 
plant of the type specimen of T. davellosum, which was 
supposed to be some cherry. 
In 1904 Milesi and Traverse published “ Saggio di una 
monografia del genere Triphragmium,” Annales Mycologici, 
II., pp. I43-I56. They keep Triphragmium davellosum and 
T. Thwaitesii distinct, but record both for Ceylon. The 
former is said to grow on Aralia nudicaulis, A. hispida, 
Paratrope terebinthacea, Hedera sp., and Amygdalaceæ in North 
America and Ceylon ; and the latter on Hedera Vahlii, H. stel- 
lata, and Heptapleurum sp. in Ceylon and Java. 
More recently Keissler has recorded ‘‘ Triphragmium davel- 
losum syn. T. Thwaitesii ” on leaves of Ahehia sp., Kandy, 
Ceylon. This is certainly a misdetermination of the host- 
plant, Ahehia not being known in Ceylon. 
No one has attacked the problem of the identity of the 
host -plant in the type specimen of T. davellosum. It seems 
to have been presumed that Berkeley’s reference to a species 
of cherry was an error ; if it was not, there is the alternative 
that the T. davellosum of 1857 was another species altogether 
different from that on Aralia. 
H. and P. Sydow (Monographia Uredinearum, III., p. 179) 
settle the difficulty by retaining T. davellosum Berk, for the 
American species, and T . Thwaitesii B. & Br. for the Ceylon 
species. They state that the two species are very close to 
one another, but T. Thwaitesii, as a rule, has less numerous but 
6(10)16 (22) 
