Abstract (1) Both the crescent and angle tests are group (2) tests; therefore the width of the sample must fee taken into account when the tear resistance is being calculated. (2) There is one exception to the above, and that is the case of knotty tears in group (2) tests. When knotty tears occur, the test becomes a direct tear test [group (1)], and the width is no longer important. (3) When knotty tears occur with group (1) tests, the width must be considered for, in reality, the test has become a group (2) test. (4) Some rubbers exhibit fibering during tearing, and this is explained by the cohesive forces between fibers reaching a limiting value before the limiting value of the breaking strength of a fiber is reached. (5) Differences in the mechanism of tearing with different fillers have been noted. (6) Practical evidence of the existence of maximum shear stress in subsurface planes of the material has been obtained, and a mathematical solution for a simple case has been given. (7) The angle tear method is a combination of tear initiation and tear propagation, and can be regarded as complementary to the crescent test. It cannot be regarded as a replacement for the crescent test. (8) With certain rubber compounds the component of stress in the direction of pull is too high in the case of the angle test-piece. (9) It is suggested that the unnicked crescent be used in preference to the angle test-piece for tear initiation measurements, as the unnicked crescent test-piece has more discriminating power. (10) The same proportion of straight and knotty tears occurs with both the crescent and angle methods. (11) A preliminary evaluation of the I.G. tear cutter indicates that the fairest comparison with other methods requires a single slit to be inserted in the outside circumference of the ring.