It is not an exaggeration to say that the data observed from the natural world is a mixture of the different kind of data group in many cases. Therefore, we should appropriately extract the component data group with the structure preserved. A variety of information processing techniques for that are researched. As for the observational data of the existence of etc., in the clinical bacillus inspection of the fishing data of the problem of an environmental electromagnetic radiation, the sound data, and living aquatic resources according to the age, the pollen dispersion data, and the medicines, the analysis of the Gauss mixture distribution is often needed. Then, the problem of the assumption of the pollen dispersion data this time and an example extracting the mixture distribution element (For an uncertain number of elements) is assumed to be an example, and the extraction of some normal distribution that is the component is tried.