The author has observed some 50 ridges in the vicinity of Arvidsjaure in northern Sweden, some of which were examined in an attempt to find out how they were formed. The locations of the ridges, their shapes, the fact that they consist of a loose sandy till as well as glaciofluvial sediments and the lack of preferred orientation of the elongated stones show that the ridges, in the opinion of the author, were formed through the accumulation in open crevasses of a thin, more or less dead ice.