This paper demonstrates that the use of quadrature to designate pseudocomponents of continuous systems in solid-fluid phase equilibrium computations can lead to results that would be inconsistent with what one would expect from actual (more continuous) descriptions. Differential distillation (liquid + vapor) and differential crystallization (liquid + solid) computations are performed using two compositional representations: (1) a 94-component integer mixture of n-paraffinic species and (2) a 12-component quadrature representation of the same n-paraffinic mixture. Results for the differential distillation process do not vary significantly between the two representations; however, corresponding results for the differential crystallization process are dramatically different. Computations suggest that strong nonideality in a mixture phase can render its pseudocomponent description inadequate, leading to unrealistic results when that phase equilibrates with another.