Delta-haemolysin has a much lower specific activity than the other haemolysins, and in low dilutions lyses a wide range of erythrocytes; it is usually assayed with human or horse erythrocytes and these have indicated as many as 300 haemolytic units per mg (HU per mg) in purified preparations (Heatley, 1971 ; Kreger et al., 1971). Heatley (1971) found that ,8-haemolysin was capable of interfering with the assay of 8-haemolysin when human erythrocytes were used. Lysis of bacterial protoplasts (Bernheimer et al., 1972) provides a more sensitive assay (610 units per mg) but is less convenient when a large number of samples is screened. The release of radioactively labelled nucleotides from tissue-culture cells (Thelestam, Mollby and Wadstrom, 1973) also provides a sensitive assay (approximately 600 units per mg purified protein) in which interference by other haemolysins does not occur. The simple haemolytic assay is still routinely used for 8-haemolysin, but a more specific and sensitive assay would be desirable. The work described here resulted from a preliminary survey of the sensitivity of marine fish erythrocytes to staphylococcal haemolysins and we describe the use of fish erythrocytes for the assay of staphylococcal 8-haemoly sin. MATERIALS AND METHODSStaphylococcal strains. Staphylococcus aureus strains NCTC10345 and NCTC7121 (Wood 46) were obtained from Professor J. P. Arbuthnott. The remainder, obtained from Dr S. McKay, comprised S. uureus strain BB, a ,fl-toxinogenic bovine-mastitis isolate, and clinical isolates of S. aureus (strains SM9, SM10, SM14, JK21) and S. epidermidis (strain SM6). Cultures were maintained on nutrient agar slopes.

The preservation of erythrocytes from cod (Gadus morhua), saithe (Pollachius virens) and mackerel (Scomber scombrus) at -196" C was studied using dimethyl sulphoxide (DMSO) as a cryoprotectant. Erythrocyte recoveries of greater than 90% were obtained from all species and cod erythrocytes were stored for eighteen months with insignificant lysis. Larger quantities of blood were stored by removal of plasma from citrated blood prior to the addition of DMSO solution, and by storage of pelleted frozen blood in aluminium canisters in liquid nitrogen. Maximum recoveries of washed intact erythrocytes required thawing of pellets in 125% DMSO solution and washing with buffer containing decreasing concentrations of DMSO. Washed erythrocytes kept at 4" for at least two days showed little haemolysis, were morphologically similar to fresh erythrocytes and equally susceptible to the 6-haemolysin of Staphylococcus aureus.

Peritoneal leucocytosis, with an increased percentage of eosinophils, was found in mice which had been infected with Schistosoma mansoni for 7 weeks or longer. Specific IgG against worm and egg antigens increased in peritoneal fluids and their corresponding sera respectively 5 and 7 weeks after infection. An intraperitoneal challenge with schistosomula elicited neutrophilia in all mice regardless of immune status, as well as infiltration of eosinophils and macrophages in infected mice. The secondary eosinophilia occurred in mice previously infected for 1 week or longer, whereas the infiltration of macrophages occurred only after worms from the primary infection had started laying eggs. Unlike the eosinophilia the macrophage response required infection with bisexual populations of cercariae. Injection of previously infected mice with Escherichia, Trichinella or Toxocara failed to increase the proportions of eosinophils and macrophages. Schistosomula-induced eosinophilia could be elicited in passively sensitized mice. Intraperitoneal injection of PBS extract of adult worms elicited eosinophilia in infected mice and neutrophilia in normal mice. Two chromatographic fractions induced eosinophilia and the third only neutrophilia. The relevance of these observations to host response to S. mansoni infections is discussed.