250 
Psyche 
[Vol. 94 
of a genus (McCluskey 1973, 1974). A clear exception is the genus 
Myrmecocystus, where the members of one subgenus are diurnal, 
and of another subgenus, nocturnal (Snelling 1976). 
The field rhythm for the E. tuberculatum colony was the same as 
for the one reported by Wheeler (1986), also on BCI, but in July and 
October, during the rainy season (though the two days were neither 
rainy nor overcast). My January observations were made during the 
transition to the dry season. For her colony there was a mass exit at 
dusk, with continued high foraging during the night, reduced to a 
very low level through the day. However, in Costa Rica in Sep- 
tember she saw much foraging in the morning as well as at night, 
with no mass exit at dusk. (The habitat was different, e.g., dry forest 
rather than moist forest as on BCI.) 
Long ago as these observations were made, I still remember them 
vividly — such a clear demonstration of circadian rhythm in the 
worker caste, not always seen in experience with other species. It 
made the strenuous around-the-clock observations worthwhile. 
Summary 
Workers from one colony each of E. tuberculatum and E. ruidum 
were studied on Barro Colorado Island, then in the laboratory. The 
peak number out near the field nest was early night for tuberculat- 
um, but during the day for ruidum. In LD the rhythms were likewise 
opposite. In DD the tuberculatum rhythm persisted. Noteworthy 
here is the apparent species difference in phase of rhythm, in both 
field and laboratory; and the persistent rhythm, not always so 
obvious in the worker caste of ants. 
Acknowledgments 
This research was supported by postdoctoral fellowship NSF 
49101 at the Biological Laboratories of Harvard University. I thank 
W. L. Brown, Jr., for his aid in determining the ant species, as well 
as in so many other ways while together on the BCI trip, and now 
for reading the manuscript; and E. O. Wilson for making all this 
possible, and for such a free use of his laboratory. 
