In the course of an investigation of the influence of bismuth on copper carried out some years ago, (1) the author found that phosphorus exerted a beneficial effect on bismuth-copper alloys of low bismuth and oxygen content by neutralizing the embrittling effect of the bismuth. In the intervening years, a further examination of bismuth-phosphorus-copper alloys has been canied out and the influence of other elements and combinations of elements on copper containing bismuth has been studied.The results recorded in the present paper are part of this investigation.MATERIALSAll the alloys were made from high-grade copper (KS.A. wire-bar) of the following composition: copper 9995%, oxygen 0,040'%, iron 00015%, sulphur less than 0-001%, bismuth less than (j>'OOl%, other elements each less than O 0010/<-. the bar was hot rolled and cold drawn to strip which was cut into short lengths and cleaned in nitric acid ready for melting.a piece of this 'strip was drawn to wire 008 in. diam., annealed at 5000 c, and measured for electrical conductivity as a check on purity. the result was 1015% of the i.e'.c. standard.in making the alloys, phosphorus was added as 14% phosphor-copper, bismuth as 17% bismuth-copper, and the other constituents either as copper alloys or as elements of a high degree of purity.preparatiqn of stripsthe charges were melted in plumbago crucibles (morgan salamander brand) in an oil-fired furnace, le;, in the manner commonly adopted in industrial practice. the copper was first melted under charcoal, well stirred with a chaned stick, and the necessary alloying elements added and stirred in with the stick. the weight of a charge varied according to requirements but generally was about 4 lb. the charges were poured into slightly tapered brass moulds heated to about 1000 c and dressed with a mixture of mineral oil and tallow. they were not only poured under charcoal through the burning mould dressing but also- through a ring of flame produced by burning kerosene on asbestos cord wrapped round the top of the mould. experiments carried out with plain copper proved that by this means oxygen could be excluded from the stream of metal. the billets were 75 in. long x o83 in. diam. x 13 lb. weight.the billets were cropped about 1 in. from the top, cleaned in nitric acid, and measured for density by the usual method of weighing in air and -in water. as these measurements were required only as a check on casting, technique a high degree of accuracy was not essential (accuracy="-+" 02%).>