IT is impossible to manufacture steel which does not contain non-metallic inclusions to a greater or less extent. These have an important effect on its properties, particularly in producing defects and causing failures to a degree which is not sufficiently realised. Mr. A. McCance, who presented a most able study of this subject at the May meeting of the Iron and Steel Institute, states that much defective steel is bad solely because of the number of non-metallic particles which it contains, and that -fully 90 per cent, of the failures due to faulty material which have come under his notice are traceable to this cause alone. He states further that when material has cracked under a stress which experience shows it should safely have carried., it is advisable to examine the crack along its whole length, and when this is done, in many cases it will be found that the crack passes through groups of inclusions, while in cases in which it can be traced to its origin it is not unusual to find that it has started from a segregation of non-metallic particles. He.treated a piece of steel in such a way as to produce slight intercrvstalline brittleness, and then stressed it above the elastic limit. A, number of small cracks appeared, and in nearly every case they started from one or more non-metallic inclusions.