Threaded code allows service routines to pass or receive data according to any convention, with parameters immediately following the threaded link. This approach can save space and time by eliminating unnecessary links. For example, a two-parameter routine on the PDP-11 can be structured with parameters directly following the threaded link. Threaded code can be combined with hard code for efficiency, especially when switching between modes is quick. In conclusion, threaded code offers an attractive alternative to hard code in certain situations, saving space with minimal time cost. The FORTRAN IV compiler for DEC's PDP-11 was developed to generate threaded code, with improvements suggested by various contributors. The paper presents efficient algorithms for graph manipulation, including partitioning into connected components, biconnected components, and simple paths. These algorithms use depth-first search and have time and space complexity proportional to max(V, E), where V is the number of vertices and E is the number of edges. The algorithms are implemented in Algol and are designed for random-access computers. They are efficient for planarity testing and other graph analysis tasks. The algorithms are described in detail, including their correctness proofs and time bounds. The paper also discusses the implementation of these algorithms and their use in graph manipulation.Threaded code allows service routines to pass or receive data according to any convention, with parameters immediately following the threaded link. This approach can save space and time by eliminating unnecessary links. For example, a two-parameter routine on the PDP-11 can be structured with parameters directly following the threaded link. Threaded code can be combined with hard code for efficiency, especially when switching between modes is quick. In conclusion, threaded code offers an attractive alternative to hard code in certain situations, saving space with minimal time cost. The FORTRAN IV compiler for DEC's PDP-11 was developed to generate threaded code, with improvements suggested by various contributors. The paper presents efficient algorithms for graph manipulation, including partitioning into connected components, biconnected components, and simple paths. These algorithms use depth-first search and have time and space complexity proportional to max(V, E), where V is the number of vertices and E is the number of edges. The algorithms are implemented in Algol and are designed for random-access computers. They are efficient for planarity testing and other graph analysis tasks. The algorithms are described in detail, including their correctness proofs and time bounds. The paper also discusses the implementation of these algorithms and their use in graph manipulation.