Some examinations are made about errors of the water equivalent of snow measured with a snow-sampler. It is found that both the change of the sampling deficiency of a sampler related to the sorts of deposited snow and the irregular variation of the snow depth in an observation field might be the main factors that disturbed the uniformity of daily successive data. The former has not been yet sufficiently studied, but it is show that the latter could be modified to some extent by comparing it with a fixed snow-tsake in snow melting season.