1975 ] Robinson and Pratt — Phenology of Hexacentrus 3 r 9 
During the day the insect responded to the movements of flies by 
orienting towards them. The impression gained was that the animal 
was responding to visual stimuli. If feeding occurs principally at 
night (when stridulation occurs), visual prey detection seems less 
likely. We do not know that the animal is active only at night. In 
captivity it certainly feeds by day and in the wild it assumes posi- 
tions, by day, on vegetation from which it could capture prey. 
The stridulation 
The males stridulate at night from the upper surface of leaves in 
secondary growth and in forest- and roadside-edge clearings. The 
stridulation is bisyllabic, “metallic,” and repeated over lcng periods 
of time. Figure 3 shows sonagraphic representation of the sound. 
It is based on a recording made at 1% inches per second on a Uher 
3000S tape recorder. Some harmonics are visible, but the frequency 
range of the principal stridulation is clearly visible in the sonagram. 
Striduating males stand with their tegmina, partly opened, i.e., 
the two tegmina are not touching along their posterior margins 
(which normally lie along the midline of the body when the insect 
is not in flight or singing). 
The song of Hexacentrus mundus has all the characteristics of a 
calling song as defined and described by Dumortier (1963). 
The census 
The census was carried out on one day per week between 2000 
and 2100 hours, between 1.5-2 hours after nightfall, irrespective of 
the weather conditions. Locations of the transects involved are 
shown diagramatically in Figure 4. They were always censused in 
the same order, I, II, and III. The technique was very simple. One 
of us simply walked along each transect to the end and back (tran- 
sects I and III) or around twice (transect II), counting all the sing- 
ing Hexacentrus that were audible on the two journeys. Individuals 
that were singing during both journeys were scored once only. From 
January to August (inclusive) the census was carried out by 
T.P. carried out the remainder of the year’s study. 
The transects differed in vegetational structure Transect I was 
along an unpaved road flanked by uncut verges (1-1.5 meters wide) 
and bordering coffee plantations for most of its length (100 meters). 
Transect II (ca. 200 meters) was around the outer perimeter of a 
garden which abutted coffee plantation around the entire length 
except for the short stretch of road shown in Figure 4. Transect III 
(100 meters) was along a footpath connecting House 4 with the 
