I mean when I deduct the Einstein’s box Bohr designed , I can not obtain the ΔT •ΔE > h like Bohr. 1) After balancing, the position of the spring-balance in weighing the box at the time 0 (T0) is Q0, and the weight of the box is M0. Because the experiment did not restrict the measuring accuracy of spring-balance position, so we suggested ΔQ≈0. 2) After rebalancing and the position of the spring-balance localized at Q2, we open the launch window at the time 1 (T1), and close it at he time 2 (T2). The total time (T) for photon to launch is equal to T2-T1; the distance the box moves in the direction of gravity is q; based on the general theory of relativity, the error of T2 we measured and the T (= T2-T1) we calculated is ΔT, and ΔT/ T = (1/c2) • g•Δq. Please pay attention, q≠ΔQ here. 3) The position of spring-balance in the second weighing of the box at time 3 (T3) is Q1, and the weight of the box is M1, the reduction weight of the box measured in the experiment, m, is equal to M0-M1. 4) At time 4 (T4), after rebalancing, the position of the spring-balance is Q2, and the actual decrease weight of the box M is equal to M1-M2. The experiment can not proved that T3-T3﹥T4 T2, which means that the second reading Q is performed after photon emission and spring-balance rebalancing. Thus we can not concluded from the experiment that Q0-Q1=Q0-Q2, that is ΔM=M-m，ΔE=ΔMc2. So how can Bohr calculate to obtain ΔT •ΔE > h?