Hering reported that the area over which he could bifixate a target was smaller at near convergence distances than far convergence distances and predicted that in extreme horizontal gaze positions, the temporally directed eye lags behind the nasally directed eye. We tested these predictions using a subjective index of eye position. Experiment 1 confirmed that the bifixation field was significantly smaller at near convergence distances. When bifixation broke down at the near distance, the nasally directed eye lagged behind the temporally directed eye for all observers. At the far distance, the nasally directed eye preceded the temporally directed eye for four of six observers. Experiment 2 also confirmed that the bifixation field was smaller at near convergence distances but the nasally directed eye always lagged behind the temporally directed eye at the limits of the bifixation field. We confirmed Hering’s first prediction that the bifixation field is smaller at near convergence distances than at far ones. However, the majority of our results indicate that the nasally directed eye lags behind the temporally directed eye at the limits of the bifixation field, contrary to Hering’s prediction. We conclude that the eyes drift toward their tonic state of vergence when fusion breaks.