The molecular weight distribution of a linear homologous polymer is usually obtained empirically for any particular sample. Sample-to-sample comparisons are made in terms of the weight- or number-average molecular weights and graphic displays of the distribution curves. Such treatment generally precludes data interpretations in which a distribution can be described in terms of differing proportions of mixed populations. However, a statistical continuous univariate distribution function such as the Weibull can be applied to the molecular weight distributions measured for cellulose nitrate samples. The size distributions of samples degraded under some gentle stress may then be indentified as consisting of differing proportions of populations with similar characteristic distributions. Analyzing data in this way should permit interpretation of the results as for breakdown of oligomers in a small-molecule system.