Dolomite-mottled limestones (and fully dolomitized equivalents) are widely distributed in epeiric successions and constitute the greater part of the Ordovician Yeoman and lower Red River Formations of southern Saskatchewan and Manitoba. In Manitoba they are exploited as building stone (Tyndall Stone) and in this rock the mottles are branching cylindrical structures, composed of dolomite and containing meniscus-filled burrows, that are set within an undolomitized biomicrite matrix. Late diagenetic (Devonian?) dolomitization of mottle-matrices has generated colour-mottled dolomites. Control of dolomitization patterns by burrows and the preferential dolomitization of aragonitic skeleta are insufficient evidence for concluding that dolomitization was penecontemporaneous. The resemblance of mottles to burrow-systems (Spongeliomorpha), and the eccentricity of contained burrows (among other evidence) indicates mottles are replaced burrow-fills, not envelopes of dolomitized sediment around burrows. The relations between mottles, contained burrows and dolomitized fossils indicate that the original sediments underwent extensive synsedimentary lithification and were further modified by multigeneration burrowing, internal sedimentation (as a consequence of burrowing activities), and early-diagenetic dissolution of skeletal aragonite. These processes created a markedly heterogeneous sediment comprising magnesian calcite (?)-cemented matrices that surrounded unlithified sediment-fills of burrows and skeletal aragonite molds. This heterogeneity was exploited by dolomitizing brines causing preferential replacement of the unlithified and more permeable parts of the sediment. Anhydrite pseudomorphs after later-diagenetic halite crystals are restricted to dolomite mottles. This restriction indicates that, when halite grew, the burrow-fills (now mottles) were still unlithified and could be displaced. This implies a persistence in time (until late diagenesis) of the sediment heterogeneity--a conclusion of importance when attempting to date the dolomitization process. Dolomitization may have occurred at the time overlying Herald (upper Red River) evaporites were deposited, by reaction of the sediments in the still-permeable Yeoman burrow-networks with refluxing magnesium-rich brines capable of precipitating halite. However, the stratigraphic distribution of mottle-dolomitization does not lend support for this interpretation. An inferred resemblance between some environmental variables in ancient epeiric carbonate-depositing environments and those of Holocene deep-water, semi-enclosed basins suggests that the Ordovician sediments may have been largely composed of magnesian calcite. It is believed that magnesium released during stabilization of these magnesian calcites allowed dolomitization of the more permeable parts of the sediment, perhaps early in the diagenetic history. Such dolomitized sediment remained unlithified to allow the late diagenetic growth of displacive halite crystals.