IN drawing up specifications of the chemical composition of carbon steels the following five elements are invariably included: carbon, sili con, manganese, phosphorus, and sulphur. With respect to the first three it is always specified that the percentage present shall fall within certain limits, whereas as regards the last two only an upper limit is demanded. These specifications are based on the assumptions that while carbon, sili con, and manganese confer desirable properties on the iron with which they are alloyed, the char acter and degree of which can be regulated by the amount introduced, phosphorus and sulphur act unfavourably, and should be kept down to the lowest possible figures. They are universally re garded as embrittling agents, which must on no account be present above a certain limit in any particular case. In fact, if it were possible to reduce the percentage of these elements to nil in commercial steels, few, if any, engineers would hesitate to specify that such steels should be absolutely free from them.