**4. Discussion**

In a demanding task with spatially varying rapid presentation, we tested the impact of early occurrence of features that were typical of the second target. The original variation of the present study from several earlier studies [3–12,15] was that lures resembled T2 in two di fferent, alternative ways. Our major expectation about these two types of lures was confirmed: Their e ffects on T2 identification grossly diverged (main e ffect of Lure Type *p* < 0.001). T2 identification was impeded by color lures and was mainly improved by digit lures.

We had assumed as a mechanism underlying this divergence of e ffects that color lures falsely attract attention to their stream while digit lures positively prime the identification of digits. From this viewpoint, the obtained results will be discussed in detail. It will be concluded that these mechanisms were indeed at work but cannot account for the whole pattern of results.
