An important challenge in studying evolutionary biology is to reconstruct the past based primarily on evidence from the present by studying the comparative biology and molecular phylogeny of present-dayorganisms. Studying the fossil record may help fill some of the gaps in our understanding. Unfortunately, the cardiovascular system does not fossilize; therefore, any rendition of the evolutionary history of the endothelium is at best speculative. In constructing a hierarchical pattern of evolution based on the comparative analysis of structure and function, it is assumed that two structures that look similar are closely related, however, structures also may look similar because they have evolved similar adaptations, so called convergent evolution. For example, the arthropod Tachypleus tridentatus (Japanese horseshoe crab) contains a clotting system consisting of a protease cascade of three serine protease zymogens and a clottable protein (coagulogen). The components of this cascade are stored in granules in hemocytes and released in response to foreign substances such as lipopolysaccharide (LPS). Hemolymph coagulation in the horseshoe crab shares many common features with the vertebrate-clotting system (cascade of serine protease zymogens, clottable protein, and LPS response), but the proteins involved are quite different and have arisen independently of each other. Conversely, structures that no longer look similar nor appear to share a common function may have arisen from a common ancestor through divergent evolution. Similar pitfalls befall the analysis of molecular sequences: Because genomes have evolved through duplication events, many coding sequences may share a degree of sequence identity with each other; however, this does not mean that they share a common function. For example, it is widely believed that large regional or whole genome duplications have contributed to the structure of vertebrate genomes (1,2).