KERN: GYMNOSPORANGIUM AND ROESTELIA 463 
any accompanying description. The name has since been used, 
but in an examination of the literature no publication with a de- 
scription has been found. A Colorado collection was later issued 
under the same name (Ellis & Ev. Fungi Columb. 1293, May, 
1898), but it is very different both in gross and microscopical ap- 
pearance and belongs with the new species proposed below. The 
geographical range of R. Harknessiana suggests a possibility of a 
connection with G. Libocedri (P. Henn.), which has the same lim- 
ited distribution, but there is at present no other clue to introduce 
as evidence of their relationship. 
Roestelia Harknessianoides sp. nov. 
entire surface, cylindrical, o. 5—0.8 mm. in diameter, 2-4 mm. high ; 
peridium yellowish-white, rather tough, not becoming lacerate, 
cells rhomboidal in longitudinal section, 4 5-55 by 65-100», over- 
lapping, outer wall moderately thick, 5-8 /#, smooth, inner wall 
very thick, 27~ 35 #4, moderately and closely verrucose with slightly 
irregular papillae ; aeciospores globoid, 23-27 by 25-31 yw, wall 
pale-yellow, 2—2.5 yw, finely verrucose, pores obscure, scattered. 
On fruit of Amelanchier oreophila A. Nels., Glenwood Springs, 
Colorado, July, 1905, Z. Bethel (type in herbarium of toa Oe 
Arthur). 
On fruit of Amelanchier alnifolia Nutt. (?), Colorado, £. 
Bethel (in Ellis & Ev. Fungi 1293). 
A very interesting species occurring only on the fruit. It has 
been confused with R. flarknessiana from which it differs in having 
shorter, lightly-colored peridia, peridial cells with a much thicker 
inner wall, and finer, closer markings, and spores with a lighter 
wall. It hasa superficial resemblance and a similar habit of growth 
tothe genuine R. Harknessiana and in separating it the name 
Harknessianoides has been chosen in order that it may carry with it 
this Suggestion. This is the species mentioned above as having a 
Possible connection with Gymnosporangium inconspicuum., 
LAFAYETTE, 
INDIANA, 
