Figure 5.25 shows instantaneous schlieren images of buoyancy-driven convection around the crystal of varying sizes (small, medium and large) over a duration of 3 hours. Compared to the time durations generally required, this is a small time scale. Figure 5.25 shows that short time transients are indeed present in the solution, but they decay within a time frame of three hours for the three sizes considered. Over this time frame, the change in the crystal size was found to be negligible. Figure 5.25 shows that the strength of convection is weaker during the transient phase, and is the most vigorous when steady state appropriate for the size of the crystal is reached. The convection plumes during the transient phase can be unsteady and unsymmetrical. The loss of symmetry is related to the imperfections in the initial crystal geometry. The unsymmetrical plume is responsible for inducing flow unsteadiness above the crystal. With the passage of time, minor imperfections of the crystal are accommodated and the plume becomes symmetric as well as steady. The schlieren images in Figure 5.25 recorded at the end of 3 hours of experimentation correlate well with those shown in Figure 5.23 (Lecture 31). The similarity of buoyant plumes in the two configurations reveals that the long-time crystal growth process is indeed quasi-steady. Figure 5.25 also shows that departure from quasi-steady growth conditions can result in loss of symmetry and unsteady plume movement, both of which are undesirable from the view-point of crystal quality.