ABSTRACT
Prior to the development and ready availability of computers the difficulty of solution of large numbers of simultaneous equations had led engineers and scientists towards the use of methods of successive approximation for the solution of their problems. These methods rely on the continuing application of a relatively simple (but limited) procedure, which gives at any instant an approximate solution to the problem investigated. Repetition of the process can increasingly refine the accuracy of calculation, for the solution converges towards the numerically correct, and indeed it is one of the merits of the process that it can be stopped at any point where further refinement of numerical accuracy is not warranted by the premises.
