THE problem of brittle fracture in structural steel has been vigorously attacked by powerful research efforts in many countries over the past sixteen years or so@ but remains in some respects intractable. The main difficulty seems to lie in defining the problem itself@ and in isolating the essential features. The investigations have been characterized by conflicts of ideas on these essentials@ and on their interpretation@ possibly because the problem has brought together into one forum@ as it were@ several branches of scientific and technical endeavour which in the past have functioned@ to a large extent@ independently. This has led to misuiiderstamlings due to differences in terminology and other difficulties due to conflicts of interest.