(a) The new map in AB is different from extended A and B maps. Top: Representative firing rate map in the divided A|B environment and the merged environment AB. Scale bar, 50 cm. Center: Simulated grid pattern A extended to the size of the merged map AB (200 x 200 cm) and the corresponding sliding correlation heat map between AB vs. extended pattern A, color-coded from dark blue to dark red (color bar). Scale bar, 50 cm. Bottom: same as center column but for the extended grid pattern of map B. (b) Left: Average sliding correlation heat map extended pattern A x AB for all cells and rats color coded as in a (n = 128 cells, 10 rats). Note that the correlation of the distal 10th percentile band of the AB map on the A side was significantly higher with extended map A (0.53 ± 0.02) than with extended map B (0.04 ± 0.02, mean ± s.e.m., n = 128 cells, 10 rats, t(254) = 17.4, P = 2.7 x 10−45, two-sided Student’s t-test). Right: same as to the left but for extended map B. The distal 10th percentile band of the AB map had a significantly higher correlation with extended map B (0.59 ± 0.01) than with extended map A (0.02 ± 0.02, mean ± s.e.m., n = 128 cells, 10 rats, t(254) = −16.7, P = 5.9 x 10−43, two-sided Student’s t-test). (c) When the central wall was reintroduced after successive trials in the merged environment, the original grid maps A and B were fully re-expressed. Left: correlation between map A on the last trial before AB and map A on the first trial after AB: 0.75 ± 0.02, mean ± s.e.m., n = 28 cells, 8 rats, t(27) = 33.7, P = 1.30 x 10−23 two-sided Student’s t-test; Right: same as to the left but for map B: 0.82 ± 0.01, mean ± s.e.m., n = 20 cells, 7 rats, t(19)=60.76, P = 13.13 x 10−23, two-sided Student’s t-test; up to 4 consecutive AB trials intervened between the two trials in the divided environment).