In the initial part of the paper we attempt to summarize on a qualitative level the factors which have been found to be important in phase separation phenomena. In particular, we focus our attention on the thermodynamics and kinetics of the spinodal and nucleation mechanisms. Continuity (or connectivity) between both decay mechanisms, as one crosses the spinodal curve, is emphasized. In the second part of the paper, spinodal and nucleation mechanisms are applied to natural magmatic fluids. This application appears of great geological interest because has significant effects on the cooling and differentiation of magmas. Fluctuations give rise to the transient appearance of clusters of different sizes, compositions, and densities whose equilibrium abundance varies according to their free energy and supersaturation. The peaks and valleys resulting in the composition waves eventually become two discrete phases. Furthermore, phenomena that are too fast to be observed in magmas can be seen easily in igneous rocks, and vice-versa. In this manner, different geological conditions (particularly cooling rates) produce different final textures in the igneous rocks. Finally, considering the ''truly primitive'' basic magma as an initially simple and homogeneous fluid characterized by density fluctuations in critical conditions, a scheme of magma differentiation is proposed.