22 
Psyche 
[Vol. 95 
(at least over a few weeks). All observations were carried out during 
the day from 8 to 16 h (Solar Time), which covered all the period of 
activity of these strictly diurnal ants (Lenoir et ah, in prep.). 
Results 
1 . Number of queens per colony 
Excavation of a nest is easy because of its predictable structure. 
Under the superficial chambers is a vertical well, which covers 
chambers where a great number of workers are packed with brood 
and the queen (Cagniant 1976b). Sometimes the well begins after a 
more or less inclined tunnel. In each case the queen was found 
(except when digging was impossible — for example when the nest 
was under a large stone). The queen was located at various depths, 
generally 40 to 80 cm deep in April. She moves up in May when the 
soil is moist, she goes deeper during drier periods. The exact posi- 
tion depends on the nature of the substrate (Cagniant 1976a). In 
sand, for example on the edge of the Etang de Leucate, nests are 
above the water level, which is at 60 cm. In limestone they are 
deeper. More than 150 nests were dug up; each contained only one 
queen. During a short period following swarming, however, colonies 
may contain several inseminated females and are temporarily 
polygynous (see below). In May 1986 one colony was found at 
St-Hippolyte (Pyr-Or.) with 3 queens. This colony was reared in the 
laboratory, and we observed the rejection of supernumerary queens 
on the 5th of July. The two rejected queens were dissected and it 
appeared that they were not inseminated, so they cannot be consid- 
ered as true queens. 
2. Colony size 
Fig. 1 represents the number of workers in late April/ beginning 
of May before the first brood appearance. It is known that no brood 
is present during wintering (Cagniant 1976b, Retana 1986). Estima- 
tion of colony size is relatively easy. The nest can be excavated 
during the resting hours when all the ants are in the nest. With the 
help of a battery-operated vacuum equipment all the workers and 
brood can be systematically removed and counted. 
The mean number of workers is 675.5 (SD 440, n = 24, range 
34-1590). Cagniant (1976b) found the mean number to be 600 
(range 5-1300) and Retana (1986) states that worker number varies 
from 150 to 1500. C. bicolor colonies have, on average, 2600 
