1953] 
Werner' — North American Epicauta 
109 
total length of antenna of 1000 units) : 89/38, 43/33, 
124/47, 92/49, 94/54, 91/50, 91/49, 82/49, 89/48, 86/43, 
118/38. Segment V is the widest and the antenna tapers 
gradually toward the apex, segments III to X being slightly 
flattened and obliquely truncated at the apex. The most 
characteristic feature is the presence of very short, ap- 
pressed, dark hairs on the underside of segments III to XI. 
These are directed perpendicularly to the axis and point 
to the midline of the antenna from each side. They can be 
seen only with fair magnification and proper illumination. 
At low magnification the surface of the underside of the 
antenna appears glabrous and roughened ; the upperside has 
moderately dense, suberect pubescence. 
The presence of these appressed hairs on the underside 
of the male antenna distinguishes this species from all 
others in our fauna, except E. ensiformis. On re-checking 
the type specimens I feel that ensiformis is based on char- 
acters that can be extremely unsafe, the absolute length of 
the antennae and the width of the segments. The ratio of 
the lengths of antennal segments seems to provide a reas- 
onably reliable criterion for species separation. But there 
is no difference in this ratio in the two. In ensiformis there 
seems to be mainly an exaggeration of the width and a 
very slight exaggeration of the length of the intermediate 
segments, to produce an ensiform antenna from the antenna 
of a normal fallax male. The types of ensiformis are like 
the fallax series in all other details. From experience with 
other species of Epicauta I am inclined to regard ensiformis 
as a synonym of fallax and fully expect that more extensive 
series will show a gradation between the two. 
what they would be if the total antennal length, made up of the total of 
the individual lengths, were 10 mm. and the figures represented 0.01 mm. 
The total of the lengths as given is 1000 3, the error resulting from 
lounding off anything under 0.01 mm. It is hoped that this system of 
recording measurements will prove superior to a segment-bA’-segment 
comparison, where errors accumulate as one progresses. It is also hoped 
that the antennae of different species will be more easily compared, since 
all would be converted to a standard 10 mm. length. The figures must 
not be taken as entire!}" diagnostic, however. They are to be considered 
as representing the antennae of a single “typical” specimen. 
