The rate of isothermal bulk crystallization of poly(chlorotrifluoroethylene), T_m =221° C, was measured from 170° to 200° C. The intrinsic bulk crystallization, which accurately followed an n = 2 law, was shown to be a result of the injection of primary nuclei sporadically in time, with one-dimensional growth of centers derived from these nuclei. The crystallites are exceedingly small. The one-dimensional growth process was isolated by nucleating specimens with seed crystals, and its temperature-dependence determined between 191° and 205° C. The seed crystal isotherms followed an n = 1 law. The temperature coefficients of the rate of nucleation and the rate of growth were both strongly negative. A theory of homogeneous nucleation that takes into account the segmental character of the polymer chains is developed in some detail. A cylindrical nucleus is assumed. In the temperature range near the melting point, region A, where the radius and length of the nucleus are unrestricted, the rate of nucleation is shown to be proportional to exp(-α/T ³ΔT ²). The nucleation rate is proportional to exp (-β/T² ΔT) in region B, which extends from somewhat below the melting point to considerably lower temperatures; the length of the nucleus has a constant value l ₀ in this region, but the radius is unrestricted. (In the above expressions, α and β are constants). Finally, at sufficiently low temperatures, region C is entered. Under certain circumstances, the rate of nucleation in region C will be extremely rapid, and correspond to a "nucleative collapse" of the supercooled liquid state. A calculation of the one-dimensional growth rate shows that it is proportional to exp(-γ/T² ΔT) where β=γ. A careful analysis of the experimental data obtained between 170° and 200° C clearly showed that both the rate of nucleation and the rate of growth were proportional to exp(-β/T ²ΔT), and not exp(-α/T ³ΔT ²). The primary nucleation event was thus of type B in this interval. A detailed analysis of the data is given, and surface free energies and the dimensions of the nuclei quoted. Quenching experiments, where the polymer was crystallized well below 170° C, gave a firm indication of the existence of region C. An experimental study was made of the extremely slow crystallization process that prevailed when the degree of crystallinity became high. The onset of this stage of the crystallization was interpreted as being the result of a massive degree of impingement. This interpretation is justified by the calculations of Lauritzen, who has given a theory of impingements that predicts a pseudoequilibrium degree of crystallinity. As indicated above, the growth process originating at homogeneous nuclei is not of a three-dimensional or spherulitic character in the region of study. Such stray spherulites as do appear in this region are shown to originate at heterogeneities. The possibility that the intrinsic growth process may become three-dimensional at crystallization temperatures sufficiently near T_m is discussed.