It has been argued that flaws exist in the standard nucleation theory of polymer crystal growth because this theory seems to predict that the notches observed in twin crystals of polyethylene should fill out so rapidly as not to be observable. However, I argue that the observed growth in the notch is reconcilable with standard nucleation theory. When proper consideration is taken for the ability of polymer chains to diffuse into the notch, the observed nucleation rates are not surprising. I consider four separate cases. In the first case, I assume that polymers, although attracted by the growth front, are not attracted strongly enough to be adsorbed prior to crystallization. In the second case, I assume that polymers adsorb readily on the crystal face, and then crystallize very near the point of first contact. In the third case, I assume that chains adsorb on the growth front prior to crystallization and that they are able to diffuse about on the growth front through large distances prior to crystallization. In the fourth case, polymers adsorb rather weakly, and reversibly, attaching and reattaching a number of times prior to crystallization. I am able to rule out the second case, while all the others predict a lower than expected nucleation rate and also predict the observed molecular weight and temperature trends.