Neuropterous Fauna of Japan. 173 
the apical portion of the posterior pair is also narrowed. 
Legs uniformly blackish-fuscous. 
Length of body (without the tails) 19 mm.; expanse 
56 mm. 
One ¢ from Kobé (Lews) in my collection. 
I have a ¢ (from Kobé) that may belong here, but it 
is scarcely probable; it has the same forms and colours for 
the head, thorax and legs; the abdomen is castaneous, 
blackish in the basal portion above ; the egg-valve very 
small, consisting of only a narrow truncated production of 
the median portion of the margin of the segment; the 
wings are browner, with the costal margin distinctly pale, 
and the apical portion of the posterior wings broader. 
I have yet three other species of Perla (each possessing 
only two ocelli) from Japan, but think it prudent not to 
describe them until more materials are before me. Neither 
of them will accord with the description of the following, 
which remains unknown to me. 
P. geniculata, Pict. 
P. geniculata, Pict., Perlides, 232, pl. 21, figs. 1, 2. 
Communicated to Pictet, by De Haan, as coming from 
Japan. 
SIALIDE. 
Nevromus, Rambur. 
NN. grandis, Thunberg. 
Hemerobius grandis, Thbg., Nov. Ins. Sp., pt. 1, 28, 
fig. 44. 
In De Selys’ collection and my own. 
Since the publication of my paper on the species of 
Chauliodes and allies, in the “ Annals and Magazine of 
Natural History” (July 1869), I have received a ¢ of the 
true Japanese NV. grandis, and am now not sure that the 
Chinese Hermes costalis and H. anticus of Walker are 
specifically the same, although these two are identical. 
Putting on one side certain colour differences, the anal 
parts have a different appearance. In the $ of NV. grandis 
the superior appendages are excised at the apex, and 
produced into a spine at the inner angle; in costalis (of 
which anticus is the ?) they appear to be shorter and 
obtuse; in the Himalayan UW. infectus, M‘Lach., they are 
extremely long, and narrowly lanceolate. 
