Experiments designed to provide quantitative measurements of nucleation rate often take advantage of the common disparity in temperatures conducive to nucleation and to growth, resulting in the popular "two-stage" technique. To date, however, there has been little quantitative work to characterize, in detail, experimental thermal histories used in two-stage treatments. In particular, although prior studies have acknowledged that the critical cluster size of interest experimentally corresponds to that encountered during the growth stage, as opposed to the nucleation stage, little attention has been paid to nucleation behaviour when a sample is heated from nucleation to growth conditions. Numerical solutions of the nucleation rate equations for the lithium disilicate system were conducted to assess the influence of thermal histories normally used in nucleation experiments on measured crystal number densities. The simulations suggest that, contrary to popular belief , the finite time required to go from nucleation to growth temperatures results not only in clusters larger than the critical size at growth conditions achieving macroscopic dimensions, but may also include a sizable fraction of smaller clusters. This effect is most pronounced when the nucleation rate is below its steady-state value.
