Static calibrations of a hot-film probe for the velocity 1 U 40cm/sec have been conducted using an annular tank and a constant-temperature anemometer system. A drift due to the change of water tempera ture could be elminated by keeping the overheating ratio at the initial value.
　A contamination drift was found to be serious in the flowing water in our wave-channel, and the correcting procesure was re-examined.
　Calibrated data of the velocity U and anemometer out-put voltage EA for the whole velocity range were applied to the usual relation EA=B+αUn, and the most suitable value of B was found to be negative.
Since, however, a linearizer is designed only for B≧O, EA cannot be perfectly linearized for the total velocity range. It was found that a good linearity was obtained by using the above relation with a positive value of B for U 6cm/sec, and that putting B=0 gave a good linearity for 1 U 12cm/sec.
　Lastly, the results of the measurements of turbulence behind a grid in the channel flow are presented.