The final report on model studies of patterned ground deals with a theory proposed for crack formation by soil desiccation. Experimental procedures and results are described in reproducing desiccation cracks under controlled conditions for soil with and without stones. The crack pattern and number of cell sides are more dependent on thickness of the soil sample and the bottom container material than temperature or humidity and show a cell area of log normal size distribution. Cracking was found to begin from the center and extend to the surface or bottom with non-uniform speed. The shape, size, porosity, and depth of scattered stones affect the initiation of cracks but do not relate to geometry and mean area of cells. Soil 'habituation' to crack in the same manner is promoted by complete soil soaking and stone admixing. Surface stones affect 'habituation' differently. Sorting of particles into desiccation cracks requires 'habituation.' This is accomplished if the soil is soaked and dried repeatedly under wind and rain without vegetation. Areas above the timber line in cold environments have the proper conditions for sorting into desiccation cracks. The unusual phenomena called 'wetting cracks, ' produced during wetting, and sorting into wetting cracks are mentioned. Factors which promote or hinder sorting into desiccation cracks are indicated. (Author).