Abstract

Summary. The division of loess deposits into warm/desert and cold/glacial is well entrenched in the literature. The nature and distribution of glacial loess are well accounted for but doubts have been cast on the existence of a true desert loess, i.e. a loess consisting of silt particles actually formed in a desert region. It appears that the loess deposits in Central Asia and North China were formed from silt particles transported out of adjacent deserts - but the particles themselves had their origins in glacial grinding and cold weathering processes in the adjacent mountains. Thus the deserts close to mountainous areas tend to have associated loess deposits. Deserts which are not particularly associated with mountains, e.g. the Sahara and Australian deserts, do not have large loess deposits nearby. Some loess deposits in the Persian/Arabian Gulf region have been observed which consist of silt-sized carbonate particles; a separate division of carbonate loess is proposed. Large scale silt formation is still one of the consequences of glacial action, however it is apparent from recent studies of quartz clasts that glacial grinding is not such an exclusive quartz silt producer as was once proposed. Much coarse-silt-sized quartz is produced directly by weathering of igneous rocks, but there appears to be little doubt that the vast majority of the silt particles for the loess deposits of the world were formed as a consequence of glacial action and cold weathering - and this includes the deposits in Central Asia and North China.