3292 FERTILIZATION OF FLOWERS BY INSECTS. 
by a few parallel ridges between which, when more highly mag- ° 
nified, transverse wavy lines, very fine and numerous, are seen. 
The only difference which these scales present compared with 
those of butterflies is, that in the former the transverse lines are 
not so fine, so regular, nor so regularly distributed over the whole 
surface ; also these lines are entirely wanting upon the scales of 
some species of Tipulariæ. Finally, while the real lepidopterous 
scales are always deeply crenate at their truncated extremity, the 
scales of gnats are not; and their truncated extremity terminates 
in a very fine margin, from which the points of the longitudinal 
ribs sometimes project. 
eg Swe a 
= 
Lr oe ee 
SS ae e 
Scales of R species of Culex. 
a, è, Suarga g rag veins of the 
Í, 
ce argins of aa wings. 
a, a (23 ct (13 si 
b, 0, HAREE proboseig ma palpi 
OSE RE aia: f the second species of Culex. 
These scales are partly i. and opaque, partly colored and transparent. It is in 
the latter that the transverse striæ frequently appear. All the figures are magnified 
400 diameters. 
I have examined several species of gnats and have found the 
proboscis, palpi, legs and abdomen clothed with scales of the same 
sort, while the thorax and the veins of the wings had forms in- 
termediate between hairs and scales. I observed this in a mag- 
nificent Tipula from Brazil given to me by my brother Fritz. 
Besides this it had as a particular ornament a long tassel of scales 
upon its legs. 
The presence of these scales upon the Tipulariz shows that 
