Measurements of wind waves have been done in the decay area of large scale flume (70m×1.5m×1.3m) including a point at the end of generating area, and one-dimensional wave spectrum have been studied associated with the analysis of wave obtained directly from the wave record. As the propagation of wind wave in decay area gradual increase of mean wave period was found associated with the decrease of wave height, and such changes of mean wave properties were closely related to the decay of wave spectrum.
　The Gaussian hypothesis of wind wave was not particularly substantiated in the generating area nor in the initial part of the decay area. However, as the propagation of the waves in decay area and as the decrease of nonlinearity of wave the frequency distribution approached gradually to the Gaussion distribution.
　When the wave propagated into the decay area, short ripples generated by wind shear damped out quickly. But carefull observation revealed the existence of tiny ripples near the crest of dominant waves even at the decay area far from the generating area. Those ripples were cinsidered to be generated by the mechanisum studid by Longuet-Higgins (1963).
　Wave spectrum in decay area showed interesting characteristics. Conspicuous energy transfer from high frequency to low frequency was observed near the dominant peak of the spectrum especially when the wave was breaking at the initial part of the decay area. The dissipation of the wave energy at hirh frequency part was very close to that due to molecular viscosity but dissipation of energy much greater than that due to molecular viscosity was observed near the dominant peak on the spectrum.
　Due to the narrow band width of wave spectum, the second peak due to monlinear effect appeared to almost highest frequency part of our spectrum. So the equilibrium range in a strict meaning could not be observed. However, the form of spectrum was fairly close to the universal form proposed by Kitaigorodski except for the excess energy concentrated near the dominant peak and the second peak at nearly twice of the frequency of dominant peak.
　Second peak has disappeard gradually as the propagation of wave in decay area. Excess energy near the dominant peak also decreased gradually in the decay area ; some of them seemed to be transferred to low frequency range by breaking of wave and some of them seemed to be dissipated by complicated mechanisum wich we have few knowledge about.