(1) FiftyS. aureusand 30 S.albusstrains obtained from infective processes in man have been investigated. Such cultural reactions as were completed did not differentiate between strains of S. albus and S. aureus, or serve to group these cocci; we found, however, that 94 per cent, of the S. aureus strains fermented mannite, as opposed to 40 per cent, of the S. albus strains. These results are in accordance with the findings of Cummins and Cumming (1913) and Hine (1922).(2) A close degree of correlation between pathogenicity for rabbits, pigment formation, and biochemical activity was noticed among theS. aureusstrains. Out of 19 pathogenic strains ofS. aureus, 9 showed medium pigmentation and 10 were strongly pigmented.(3) Anti-sera were prepared by inoculating rabbits with living cultures, 24 hours old, of the various strains. TheS. aureusrabbits and a few of theS. albusreceived two preliminary doses of the specific vaccine, but little or no protective action was afforded from the doses of living cocci injected.(4) Nineteen out of 25 cultures ofS. aureuskilled rabbits, but only 3 out of 18 cultures ofS. albus.(5) Fifteen strains ofS. aureusand 15 ofS. albuswere used for the preparation of anti-sera, and tested against as many antigens as possible, by the methods described by Dudgeon and Bamforth.(6) Cultures ofS. aureusproduced efficient anti-sera more readily thanS. albuscultures, but when an activeS. albusanti-serum was produced, it appeared to be as effective as theS. aureusanti-sera.(7) A marked difference in efficiency betweenS. aureusandS. albusantigens was noted.S. aureusantigens were almost three times more active thanS. albusantigens.(8) Pathogenic strains ofS. aureusandS. albusproduced more efficient precipitin antigens than non-pathogenic.(9) From the microscopical examination of the deposit formed in the precipitin reactions, a difference was noted betweenS. aureusandS. albusantigens.(10) The study of the precipitin reactions has failed to show any hard and fast dividing line between staphylococci of different colours, or of different cultural reactions. We are inclined to agree with Winslow, Rothberg and Parsons (1920) and with Dudgeon's original view in 1908, that the pyogenic, Gram-positive staphylococci must be regarded as members of one common family.