Summary. The Kangerdlugssuaqintrusion (68' 2r N. 32' 30" W., roughly circular and 33 km across) is of Tertiary age, like the nearby Skaergaard intrusion. At gh~ present level of erosion the outer part is of nordmarkite and inwards there are successive rings of transitional pulaskite (without quartz or nepheline), main pulaskite and, in the centre, foyaite. The different rock types are due to gradual ehanges in the amounts of cer~aia of the minerals and there are no sharp junctions such as would be expected to result from successive injections.Zones of inclusions of basalt, fallen from above, dip inwards at 30 to 60 ~ and, in places, there is a platy parallelism of the tabular alkali feldspars dipping similarly inwards. The three-dimensional form of the intrusion is pictured as resembling a pile of saucers of decreasing size, the secluence of formation being from the outer saucer of nordmarkite to the inner, and smaller saucers of increa~ingly nephelinerich syenites. Bottom accumulation of crystals is postulated to explain the disposition of the rock types, but the cause of the succession from quartz-bearing to felspathoidal syenites remains uncertain. N EPHELINE-I:earing syenites were found by the writer durirg the Watkins Expedition of 1930, on the west side of Kangerdlugssu~q fjord in east Greenland, about 20 miles from the layered basic complex which has since been named the Skaergaard intrusion. In 1932, accompanied by his brother, a further examination of the region was made and it was established that the nepheline-bearing rocks formed part of a large intrusion exhibiting a sequence of syenites from nordmarkites to pulaskites. In 1935-36, with the collaboration of W. A. Deer, the intrusion was mapped (1937, pp. 401 and 408) and shown to be a series of rings of considerable regularity, the outer being of nordmarkite which passes gradually into syenite, free from quartz, and then into syenite with increasing amounts of felspathoids, until an innermost foyaite is reached. Because of the gra.dualness of these changes, and the disposition of the various rock types, it was considered, at that time, that the change from a quartz syenite to an undersaturated nephelinesodalite syenite was probably the result of crystal fraetionation.Much preliminary loJboratory work wr~s done in the 1940's by the present writer and by Professor W. A. Deer who analysed some twenty