These tests were made to develop a satisfactory instrument for recording torsional vibrations and to make possible calculation of frequency of vibration from drawings. A very satisfactory instrument has been devised for recording torsional vibration. Many tests have been made and vibration records taken on engines of all makes and types. These records are presented and discussed. The method of calculating frequency of vibration is presented in its simplest form, and results of calculations with corresponding observed frequencies are tabulated.