Fine bare wire thermocouples (12.5 and 25 μm) embedded in polymer films have been used to generate cooling curves in a series of studies, reported earlier, involving crystallization at cooling rates as high as 2000°/min and comparable to those used in film manufacture in industry. The rapid cooling method has resulted in the extension of basic studies of crystallization to supercoolings previously inaccessible using existing methods. The results have been very enlightening and have verified existing theories, but have also generated unique information and heretofore unobserved significant changes in behavior of several polymers. The most startling result is the conclusion that the growth face of a crystal is at a higher temperature than any temperature that can be measured, thus raising serious questions regarding the interpretation of experimental data. The current study investigates the effects of embedding thermocouples in thin films of linear polyethylene on the nucleation and growth rates of spherulites in close proximity. Row nucleation is observed along the surface of the thermocouple, with lamellae growing preferentially perpendicular to the thermocouple axis. When growth rates of spherulites are plotted against spatial distance from the thermocouple, a sigmoidal dependence is observed. The distance at which the growth rate becomes independent of proximity to the thermocouple, or in other words, the width of the `characteristic zone of influence' is found to be a function of thermocouple size and cooling rate. The most remarkable finding is that spherulites growing within this characteristic zone displayed asymmetry in growth rates with radial direction. Consequently, the spherulites exhibit ellipsoidal or irregular shapes and orientation depending on the applied cooling rate. Two independent, albeit simultaneously occurring phenomena are identified to explain the observed behavior: (i) induced mechanical stress, that affects the overall orientation of the spherulites with respect to the thermocouple axis, and (ii) thermal gradient introduced by the heat conduction effect of the thermocouple. However, the relative contribution of these two factors is a function of the applied cooling rate; the behavior being dominated by stress under rapid cooling conditions, whereas the thermal gradient governed the behavior in slow cooling.