Keywords: 
• fine-grain loess /
• pretreatment method /
• ultrasonic dispersion /
• grain-size measurement

Abstract: The common pretreatment method has caused divergent results of grain size measurement of the fine loess samples. In order to understand effects of pretreated method on the fine-grain loess grain-size distribution, samples from three sits, which are located in Luochuan of central Chinese Loess Plateau, Lushi of southern margin of the Loess Plateau and Shangbaichuan of eastern Qingling Mountains in central China, were collected and pretreated under different ultrasonic dispersion conditions. The results indicate that time of ultrasonic disaggregating by the laser grain-size equipment just before measuring is very important for granulometry of the fine loess samples. On the basis of the different experiments, we recommend a new pretreatment method for the grain-size measurement for fine-grain loess samples: over 2 minutes are required for dispersing the samples by ultrasonic of the laser grain-size equipment just before measuring. This step is neglected in the previous measurement so that the divergent results are presented in granulometry of the fine loess samples. In addition, samples from a loess-paleosol section at Qiaojiayao of the southeast Loess Plateau are measured by the traditional and the new methods, respectively. The results show that our recommended pretreatment method is more reliable to obtain grain-size distribution of the fine-grain loess deposit than that of the older one.