A different kind of magmatic segregation involves liquid immiscibility. A cooling magma will sometimes precipitate droplets of a second magma that has an entirely different composition. Like oil and water, the two magmas will not mix (i.e., they are immiscible). The chemical principle governing precipitation of an immiscible liquid is the same as that governing crystallization of a mineral from a magma: when the concentration of a particular mineral within a parent magma reaches saturation, precipitation occurs. If saturation is reached at a temperature above the melting point of the mineral, a drop of liquid precipitates instead of a mineral grain. The composition of this immiscible drop is not exactly that of the pure mineral, because the liquid tends to scavenge and concentrate many elements from the parent magma, and this process can lead to rich ore deposits.

Iron sulfide is the principal constituent of most immiscible magmas, and the metals scavenged by iron sulfide liquid are copper, nickel, and the platinum group. Immiscible sulfide drops can become segregated and form immiscible magma layers in a magma chamber in the same way that cumulus layers form; then, when layers of sulfide magma cool and crystallize, the result is a deposit of ore minerals of copper, nickel, and platinum-group metals in a gangue of an iron sulfide mineral. Among the ore deposits of the world formed in this way are the Merensky Reef of the Bushveld Complex, producer of a major fraction of the world’s platinum-group metals; the Stillwater Complex, Montana, host to platinum-group deposits similar to the Merensky Reef; and the Norilsk deposits of Russia, containing large reserves of platinum-group metals.

Under suitable conditions, immiscible sulfide liquids can also become segregated from flowing lavas. An example is offered by the Kambalda nickel deposit in Western Australia. At Kambalda a nickel ore mineral, pentlandite, together with valuable by-product minerals of copper and platinum-group metals, crystallized in an iron-sulfide-rich gangue from a sulfide liquid that had become segregated from a magnesium-rich lava called a komatiite (named for the Komati River in South Africa).

A group of unique deposits formed by immiscible sulfide liquids is the Sudbury Igneous Complex in Ontario, Canada, which formed about 1.85 billion years ago. Elliptical in outline (approximately 60 kilometres long by 28 kilometres wide), the complex has the shape of a funnel pointing down into the Earth. A continuous lower zone of a mafic rock called norite lies above a discontinuous zone of gabbro, some of which contains numerous broken fragments of the underlying basement rocks and some of which is rich in sulfide ore minerals of nickel, copper, and platinum-group metals. Many hypotheses have been suggested for the origin of the Sudbury Complex. Some consider it to be an intrusive igneous complex, and some consider it a combination of intrusive and extrusive igneous rocks, but the most widely held opinion derives from the work of the American scientist Robert S. Dietz, who in 1964 suggested that the Sudbury structure is an astrobleme, the site of a large meteorite impact. The complex’s sulfide ore bodies are thought to be derived from immiscible magmas formed in Earth’s mantle as a result of the impact (and possibly mixed with meteorite material), while the uppermost layers are thought to be rock debris remaining from the impact.

A final and highly controversial group of deposits deserves mention because some geologists believe them to have formed from immiscible melts. These are magnetite deposits associated with volcanic rocks of dioritic affinity (i.e., igneous rocks intermediate in composition between granites and gabbros). There is no doubt that lava flows account for the presence of magnetite deposits in northern Chile, but there is great conjecture over whether magnetite bodies associated with these lavas formed as a result of a dioritic magma precipitating an immiscible oxide magma, as a magnetite lava formed through the melting of a previously formed sedimentary iron deposit, or as the source of a hydrothermal solution that deposited the magnetite. Many experts draw the latter conclusion. Considerable controversy also surrounds the origin of the famous Swedish iron ores at Kiruna and Gällivare. These magnetite-apatite bodies encased in volcanic rocks have been variously interpreted as having formed as immiscible oxide magmas, as iron-rich sediments that were subsequently metamorphosed, and as deposits arising from volcanic exhalations.