The size of fragments reaches 10 cm. The largest fragments are well rounded although angular specimens also occur. The occasional small boulders and pebbles impart to the marble the habit of a pudding. The puddings form lenses, but at the same time are confined to a strictly definite stratigraphic horizon in the profile of the Upper Archean. These lenses are a few meters thick and a few hundreds of meters across.
The great age of these deposits and their Archean dating are unquestion able. The entire rock complex containing the puddings was intensively metamorphosed by an intrusion of alaskite granites, whose absolute age is estimated as 2000 to 2300 million years according to the most recent data. In character these formations resemble the "marine tillites" which have been repeatedly described by A. N. Churakov in the Upper Proterozoic of the Altai-Sayan highlands and which have been known for a long time in South Africa, including the Damar system already mentioned. The mode of occurrence of the puddings amidst carbonate (marine) deposits suggests that they have been formed by the unloading of elastic material embedded in ice floes. The latter were probably formed near the rocky coasts of the continents (or islands) and were carried out into the open sea by the currents.
Proterozoic glacial formations have been somewhat better studied than the Archean ones, yet their stratigraphy is still very little known. The Proterozoic glaciation occurred on the most tremendous scale on the Canadian Shield (Huronian). Probable analogues of the Huronian tillites have been indicated in Ireland, Finland, the southern Urals (puddings of the Barangul Mountain), etc. Possibly they also include the supposedly Proterozoic glacial formations in Africa (the Nabib Desert), Brazil, New South Wales, etc.
Glacial deposits of Late Pre-Cambrian or Eocambrian are well known and have already been described in the specialized literature long ago. The majority of investigators consider the presence of glacial facies to be symptomatic of the Late Pre-Cambrian. However, the correlation of these formations at the present time encounters certain difficulties, whereas it was formerly considered to be fairly simple. As already noted, the volume of Late Pre-Cambrian which was differentiated under the name of the Sinian complex, Riphean group, and Lipalian system has been enormously enlarged by the latest absolute geochronological data, which, however, markedly contradict the general geological data. Nevertheless,
a revision of the stratigraphy of glacial formations related to Late Pre Cambrian is now indicated.
The following are the principal and best known locations manifesting the largest glaciations, beginning with the Late Pre-Cambrian.
Late Pre-Ca mb r i an : North East Land, Spitsbergen, Norway (sparagmites), Scotland (Torridonian), Podolia (Mogilev arkoses), Greenland, North America (tillites of Keweenaw and Great Salt Lake), South Urals, Yenisei Range, Patom Plateau (Figure 11), China (Nangtu tillite, etc.) India, Central Asia, South Africa (Numis tillite, etc.), Australia, New Zealand, etc.
Ordovician : Central Urals, south Germany (?), England (?), Canada, Alaska. Remarkable glacial-marine Tremadocian deposits are known in South America (the Andes), described by Harrington and Keidel.
Upper Paleozoic (Gondwana complex) : the well-known glacial deposits of Africa, South America, India, and Australia. The northern hemisphere contains insignificant manifestations of local glaciations, probably of the mountain type, such as on the western slope of the southern Urals.
Quaternary: the tremendous glaciation which comprised vast expanses in Europe, North America, and Asia. The first signs of cooling were observed in the Middle Pliocene. Older, (Paleogenic) glaciations were probably of a purely local character.
My previous works (1956, 1957) provided a brief summary of absolute geochronological data and advanced the conclusion concerning the average duration of the great glaciations, approximately 200 million years, i. e., corresponding to the galactic year. Although certain corrections have been introduced in the geochronological scale in recent years, they do not affect this general conclusion.
It is probably premature to discuss the immediate causes of the great glaciations. It must be acknowledged that the earlier hypothesis concerning the variation in the intensity of solar radiation due to the translocation of the Sun from the region of the galactic nucleus (gravitational maximum) to the peripheral regions of the Galaxy with its minimum stellar density (Lungersgauzen, 1956) is not the only possible one.