Photomicrograph of a thin section showing a fracture through a quartz sand grain and the surrounding silty dolomitic limestone matrix of an Ordovician limestone from Pennsylvania. A few dolomite crystals in the quartz grain (see right image) document that the fracture split the quartz grain as well as the carbonate, because the dolomite crystals would only be present if there had been a fracture filled in part by them and mostly by quartz that is syntaxial with the grain itself.

     This illustrates the criticality of substrates and nucleation to cementation. Where a substrate provided one pre-existing quartz crystal, quartz grew as the infillng mineral. Where the substrate provided mutiple dolomite crystals, dolomite precipitated and quartz did not. It appears that a few new dolomite crystals nucleated along the fracture in the quartz, although they could have grown on small pieces of dolomite that fell into the fracture. The implication is that substrates and nucleation can control the mineralogy of precipitating minerals. Thus the inferences about diagenetic fluids and diagenetic environments that carbonate petrologists make from cements have to be tempered by the recognition that the presence or absence of a given cement type may have been controlled by the substrate. The fluid precipitating minerals in this fracture was supersaturated with respect to dolomite and quartz; we know that it was supersaturated with respect to both only because both were present as substrates in the pre-exisitng rock.

.