1987] 
McCluskey — Circadian rhythm 
247 
Fig. 1. Above-ground field rhythm on BCI. E. tuberculatum, ants on vine within 
2 m of nest; mean ± SE, n = 3-4 (i.e., each mean is based on 3-4 days as replicates for 
given hour on graph), Jan. 10-11 and 17-21; P < .01 for difference among hours 
(one-way ANOVA). E. ruidum, 2 nests, ants within .5 m of nest. N = 2-4 (nest 
and/or day replicates), Jan. 10-11 and 17-19; P = . 01. 
arena, always in contrast with the high number for tuberculatum; 
and the record is not shown. 
E. tuberculatum was checked in two ways, DD following LD, and 
DD following light exposure at night. Fig. 3 serves to illustrate 
both, because the patterns were similar relative to what had been the 
light period. In DD (Feb. 9-11) following LD the arena count 
resembled that in LD in that it was low near what had been the 
light-on hour and high near what had been light-off. The data are 
too few to compute an accurate free-running period; but for the 
total 3-day span of counts a cosine least-squares fit (Halberg et al. 
1972) to a trial period of 24 hours is good (P < .01), thus suggesting 
a persistent rhythm. And the counts for the various times of day 
differ consistently (using the 3 days as replicates, ANOVA P < 
.001), again suggesting persistence. 
If rhythmicity is truly endogenous it should still appear after 
reversal of day-night phase. So the lights were turned on at 2000 Feb 
13 for 12 hours (i.e., through the night), then left off continuously 
