ON FUNGI PARASITIC ON SCALE-INSECTS FOUND IN FORMOSA. 
77 
mycelium 3.5-8// in diameter, and provided with mycelial hypothallus of 
grayish white color, forming a thin membrane adhering to the leaf surface 
and extending 1-2.5 mm. beyond the stroma. Pycnidia immersed in the 
stroma, irregular in shape, and opening by small round or elliptical pores or 
slits at the surface; conidiophores filiform, much branched, continuous, densely 
packed together, 55-70// x 1 // ; paraphyses similar to the conidiophore in shape, 
projecting beyond them, 85-108// long; stylospores fusiform, continuous, 
mucilaginous, hyaline, 11-13// x 1-1.5//, often oozing out in ochraceous spore- 
masses. 
Hab. — On Parlatoria zizyphi (Lucas) Sign, infesting the leaf of Citrus 
nobilis Lour. 
Formosa-. Seira, Kagi. Nov. 13, 1909. (K. Saw ad a, Nov. 13, 1909). 
Distrib. North America, Cuba and Japan. 
Remarks. Our fungus corresponds so closely in almost all important 
characters with the descriptions and figures of Aschersonia Aleyrodis of North 
America, that we are led to think it more appropriate to consider them for 
the present as one and the same rather than to treat ours as a distinct 
species. 
In the Formosan form the paraphyses are also always present. They 
are very delicately filiform, and continuous with dense homogeneous refringent 
contents. In many of these paraphyses, we noticed interspersed here and 
there portions devoid of the contents, which appear under a microscope as 
darkened sections. (PI. VI. fig. 4). Some of the paraphyses are seen to have 
lost almost all of their refringent contents. These vacant spaces sometimes 
collapse giving to the filament an appearance of a series of short cells. What 
Webber (31) considered as characteristic darkened cells is no more than the 
vacant sections in a filiform cell. The conidiophores are not simple but 
irregularly dichotomous or trichotomous, and their ultimate branchlets are 
subulate and 10-26// in length. (PI. VI. fig. 5,6). 
The color of the spore-mass in our dried specimens is generally ochracc- 
ous. As w T e have not yet examined fresh material with sufficient care, we 
could not say that the mass presents a conspicuous coral-red or rufus color as 
described by Webber and others. 
