Saliva inoculated by sandfly females during feeding stimulated production of high levels of anti-saliva antibodies. To determine whether 3 species of the genus Phlebotomus have species-specific salivary antigens we performed dot-blots and immunoblots using sera from mice, hamsters and rabbits repeatedly bitten by sandflies. Important differences were found in the antigen components of the salivary gland lysates (SGL) of Phlebotomus papatasi, P. perniciosus and P. halepensis. In total 4–9 species-specific antigens were detected in each species by immunoblotting. Cross-reactivity was not detected between P. papatasi and the other species tested; in the SGL of P. papatasi sera from animals bitten by this species recognized 5–7 major antigens while sera from animals bitten by other species did not react. A weak cross-reaction was observed between P. perniciosus and P. halepensis; in SGL from P. perniciosus, the sera from rabbits and hamsters bitten by this species recognized about 8 intense bands while sera from animals bitten by P. halepensis reacted weakly with up to 4 saliva polypeptides.

Promastigotes of Leishmania major were frequently detected in the urine droplets discharged by infected Phlebotomus papatasi and P. duboscqi females during feeding. Parasites were present in the urine of 37·5% P. papatasi and 16·1% P. duboscqi females, even in those with low intensity gut infections. Free-swimming forms (elongated nectomonads, short slender promastigotes and metacyclic forms) predominated in excreted droplets. Viability of excreted parasites was proved by cultivation on blood agar, and the presence of metacyclic forms in urine droplets was confirmed by specific fluorescence assay with 3F12 antibodies. While the release of promatigotes from the anus of the sandfly was frequent, these were rarely egested from the mouth-parts of sandfly females (1·3% for P. duboscqi and 0% for P. papatasi) fed on microcapillaries, even if the females were heavily infected. The possible role and significance of the discharge of parasites in sandfly urine are discussed.

—In order to evaluate the effect of permethrin-treated wall cloth against phlebotomine sandflies, a study lasting 26 months (February 1989 to March 1991) was conducted in Marigat Location of Baringo District, Kenya, an endemic focus of both visceral and cutaneous leishmaniases. In six experimental villages, a total of 2000 houses were fitted with cloth made of cotton and impregnated with permethrin EC 20% at a dosage of 0.5 g/m2 diluted in water. Houses in the two control villages were not fitted with cloths and were located 1.5 km away from the experimental area. Traps made of clear polythene sheeting were coated with castor oil and placed inside houses to collect endophilic sandflies. In the experimental villages, the permethrin-treated wall cloth (called Mbu cloth) reduced the population of sandflies found inside houses by 52–73%. Phlebotomus martini and P. duboscqi, vectors of visceral and cutaneous leishmaniases, were reduced by 76 and 85% respectively. In the experimental villages, highly significant differences were observed between the average numbers of sandflies collected before deployment of the treated wall cloth and those observed after thethird treatment of the cloth. No major sandfly reduction was observed however, in the control villages. There was also evidence of feeding reduction due to the permethrin.