A previous chapter in this series (1) described, primarily, the physical mapping of the 100 Mb Caenorhabditis elegans genome by fingerprinting of cosmid clones, and the linking of the contigs thus derived by YAC hybridization. At that time, the primary function of the map was to enhance the molecular genetics of the organism by facilitating the cloning of known genes, and to serve as an archive for genomic information. However, a clonal physical map - even with good alignment to the genetic map - carries only a tiny proportion of the information present in the genome. Consequently, the current objective of the C. elegans genome project (2) is to establish of the entire genomic sequence. The bacterial clone map, although incomplete by virtue of the uncloneability of regions of the genome in cosmid vectors (a factor which we shall discuss later in this chapter), has proved a sound basis for the systematic sequence analysis. The sevenfold cosmid coverage has a resolution sufficient to enable the selection of a subset of cosmids for sequencing such that, on average, each clone contributes 30 kb of unique sequence to the whole. Sequencing projects based on bacterial clone maps (3-5) of a number of other genomes of a range of sizes are also well advanced, in particular Saccharomyces cerevisiae (15 Mb; complete), Schizosaccharomyces pombe (15Mb), and Drosohpila melanogaster (150 Mb). Although it has recently been demonstrated that small bacterial genomes can be sequenced by direct shotgun sequence analysis of the entire genome with no prior mapping (6), the ability to interrelate and map clone sets, whether derived by random selection of in a directed manner, is still the most convenient route to the sequence analysis of larger genomes.