It has been shown in the last two reports that regular monochromatic waves propagating along a long array of cylinders fixed in the waves with equal distance are : (1) decayed monotonously along the array from the head of the array toward the end of the array when the wave frequency is relatively high. (2) enhanced as they propagate toward the end of the array when the wave frequency is relatively low. It has also been shown that these two phenomena can be predicted fairly accurately by the linear potential theory. On the other hand, it was shown in the second report that there exists a certain intermediate frequency range where large surface elevations are predicted in the theoretical calculations whereas actual surface elevations do not become that large. It has been suspected that this discrepancy between the experimental results and the theoretical predictions, which are based on the linear potential theory, may be attributed to the wave breaking caused by excessive surface elevations. If this is the case, measured results should become closer to those predicted by the linear potential theory as the wave-height of the incident waves is decreased. In order to examine the propriety of this assumption, experiments are reconducted in the present study while varying the wave-height of the incident waves. Experiments in irregular waves are also conducted to examine the assumption from a different point of view. An attempt is also made to modify the linear potential theory so that the discrepancies between the measured data and the theoretical predictions can be accounted for.