Developmental systems regulate the expression of phenotypic variation (Laland et al., 2015) through various types of morphological homeostasis or canalization (per Canon, 1932; Waddington, 1957). This variation is one of the main drivers of evolution by means of natural selection (Darwin, 1859). The processes of environmental and/or genetic canalization—which control the expression of symmetric phenotypic variation among individuals in a population—are known to serve as a constraint on evolution (see reviews by see reviews by Møller and Swaddle, 1997; Polak, 2003; Graham et al., 2010; Klingenberg, 2015; Webster, 2019 and references therein), whether and how homeorhesis—the process regulating within individual left-ride symmetry of individual traits—relates to phenotypic evolution is not well known; nor is the degree to which genetic and environmental canalization relate to homeorhesis. Empirical studies of homeorhesis have almost exclusively been conducted by neontologists. However, to understand the macroevolutionary consequences of morphological homeostasis, and to determine the timescale over which morphological homeostasis limits the expression of phenotypic variation—both within a species and across clades—it is necessary to study homeostasis within the fossil record (see discussion by Webster, 2019). The following studies build upon the central idea of understanding the components of variation within the Middle Devonian trilobite genus Eldredgeops. Chapter 1 presents a phylogenetic framework for the clade of interest, focusing on the taxonomic relationships of Eldredgeops within the larger context of the family Phacopidae. Using Bayesian inference and parsimony-based methods, trees were recovered that largely converged on similar topologies suggesting a greater degree of phylogenetic signal than previous phylogenetic studies examining the Phacopidae. Chapter 2 explores the relationship between the components of variation in related to eye-lens distributions in the eyes of two species of Eldredgeops. This study is the first to use a maximum likelihood method proposed by Young (2007) to estimate the symmetric and asymmetric components of variation. The final chapter, Chapter 3, further explores developmental asymmetries in the eye, as well as the relative addition of eye-lenses through ontogeny, to propose a model of eye development that can be applied to all trilobites. These studies represent the first steps in testing developmental hypotheses in deep time ultimately aimed at understanding what role developmental asymmetries may have on the evolutionary trajectory of traits.