875 and follicle-stimulating hormone (F.S.H.). These act on the testes to stimulate testosterone production (L.H. effect) and probably also some tubular activity (F.S.H. effect). The testo- sterone, if in adequate quantity, will stimulate testosterone- dependent processes. Using L.H.-R.H. to induce testicular des- cent implies that the descent is testosterone-dependent. If the tissues are incapable of responding to testosterone completely or partially (as in testicular feminisation), there will be no effect. Thus before using L.H.-R.H. therapeutically, it is impor- tant to establish that the pituitary gland can respond to it, that the testes can respond to L.H., and that the target tissue cannot respond to testosterone. Dr Illig and his colleagues (Sept. 10, p. 518) were less than scientific in their exploration of L.H.-R.H. as a "new drug". L.H.-R.H. could be the correct drug for the correct patient and it is the investigator’s role to determine the correct patient. University Department of Child Health, Royal Hospital for Sick Children, Yorkhill, Glasgow G3 8SJ WILLIAM HAMILTON STAPHYLOCOCCUS SAPROPHYTICUS INFECTIONS SIR,-Linda Pead and Dr Maskell have found (Sept. 10, p. 565) an increased recovery-rate of novobiocin-resistant, coagu- lase-negative staphylococci when they used a selective culture medium. At the 4th congress of the Medical Women’s Interna- tional Association for the Northern European Countries, held in Stockholm in June, we reported similar experience with tryptone broth containing 5 p.g novobiocin and 150 ug nali- dixic acid/ml. We studied 120 urethral specimens from fema- les ; 14 novobiocin-resistant, coagulase-negative staphylococcal strains were isolated, while only 2 such strains were detected when we inoculated the same specimens directly onto blood agar plates. 7 of the 14 strains were Staph. saprophyticus (mic- rococcus, subgroup 3), the remainder being Staph. cohnii and Staph. xylosus.1 Staph. saprophyticus does seldom occur in voided urine spe- cimens as a contaminant. Also when found in such specimens in numbers <104/ml, the finding seems generally to be of clini- cal significance. When we added novobiocin (2 g/ml) and nalidixic acid (150 g/ml) directly to voided urine specimens and incubated them overnight at 37&deg;C before seeding on blood agar we also recovered Staph. saprophyticus more often than when we cul- tured the same specimens conventionally. We studied 56 voided urine specimens containing < 104 bacteria/ml, as demonstrated by culture on blood agar plates using the cali- brated loop technique. Novobiocin-resistant staphylococci were isolated from 6 of the specimens; 3 were Staph. saprophyticus. These strains were not isolated in conventional urine cultures. Of another 250 voided urine specimens found to contain < 104 staphylococci/ml only 7 yielded strains of Staph. saprophyticus on blood agar, although many staphylococcal colonies were classified. When Staph. saprophyticus occurs in voided urine specimens in numbers >10’* organisms/ml bacterial diagnosis was no problem. Staph. saprophyticus is a common cause of acute cystour- ethritis in young women.2,3 We find that about half of all patients with urinary-tract infections caused by this organism have symptoms consistent with an upper-urinary-tract infec- tion. These infections are, however, often overlooked because the patients’ general condition is less affected than when gram- negative rods are the culprit. The patients are generally only subfebrile, but they do have back or loin pain and usually ten- 1 Subcommittee on Taxonomy of Staphylococci and Micrococci, Int. J. syst. Bact. 1976, 26, 332. 2 Maskell, R. Lancet, 1974, i, 1155. 3 Sellin, M., Cooke, D. I., Gillespie, W. A., Sylvester, D. G. A., Anderson, J. D. ibid. 1975, ii, 570. 4 Hovelius, B., M&acirc;rdh, P.-A., Petersson, A. C., Hovelius, K., Nilsson, P. O. Sv. vet Tidn. 1977, 29, 635. 5 Hovelius, B., M&acirc;rdh, P.-A., Acta path. microbiol. scand. B (in the press). derness over one or both kidneys. Urine-concentrating capa- city, as measured by the pitressin-tannate test, may be reduced. We have found Staph. saprophyticus in the skin flora of many domestic’ animals and have also isolated the organism from wounds of people in contact with animals.’ Pus from hand wounds yielded novobiocin-resistant, coagulase-negative staphylococci significantly more often when culture was done after enrichment in the selective broth than when it was done by direct inoculation onto blood agar. Urinary-tract infections caused by Staph. saprophyticus may be missed for several reasons. The bacterium grows only scantily on MacConkey’s agar; it does not reduce nitrate; and bladder-incubated voided urine specimens of patients with in- fections with Staph. saprophyticus above the bladder neck, proved by suprapubic bladder aspiration, sometimes contain only 10-10" bacteria/ml. Institute of Medical Microbiology, University of Lund, S-223 62 Lund, Sweden PER-ANDERS M&Aring;RDH BIRGITTA HOVELIUS IMMUNITY TO PATHOGENIC FREE-LIVING AM&OElig;B&AElig; SIR,-Pathogenic free-living amoebae of the genera Nagleria and Acanthamceba (the agents causing primary ammbic men- ingoencephalitis) are readily isolated from such diverse environ- ments as nasal and throat cavities, chlorinated swimming and potable waters, soil, sewage, thermal pools, and temperate lakes and rivers,’-4 yet the disease is rare; only about 90 cases were recorded in Willaert’s review.s With the possibility of host-related susceptibility factors’&middot;6-a in mind we did a serolo- gical survey of healthy New Zealanders to test for antibodies against free-living amoebae as judged by the indirect fluores- cent-antibody test. Antibodies were found in all of the 200 un- diluted sera. Titres ranged from neat to 1/120 for pathogenic and non-pathogenic Ncegleria and from neat to 1/80 for patho- genic and non-pathogenic Acanthamaeba. Antibodies to these amoebae in healthy populations have been reported by others.9-11 Concurrent with the survey, we did random neutral- isation tests using hyperimmune rabbit and normal positive human sera in Vero cell cultures. No neutralisation was obtained with either unheated or heated (56&deg;C/30 min) hyper- immune rabbit or normal human sera for N. fowleri. Only un- heated adult sera, as opposed to heated or cord sera, neutral- ised A. culbertsoni at a titre of 1/10 to 1/20. The addition of complement did not affect these results, nor did the use of hyperimmune rabbit anti-A. culbertsoni sera with a titre of 1/1000. In view of the failure of the hyperimmune serum to neutralise A. culbertsoni, we postulated that normal human serum contains some heat-labile natural antibody. The neutral- ising action of this natural antibody may explain why, despite the prevalence of pathogenic Acanthamceba in the environ- ment’ few cases have been reported, most of them being in defence-weakened patients2 or in parts of the body with low immunity.12 Because of the antigenic cross-reactivity between non-patho- genic and pathogenic Ncegleria (and Acanthamaeba13) and the apparent failure of humoral antibodies to neutralise N. fowleri the cell-mediated immune system was examined in guineapigs by the macrophage-inhibition test and delayed-hypersensitivity 1. Culbertson, C. G. A. Rev. Microbiol. 1971, 25, 231. 2 Chang, S L. Crit. Rev. Microbiol. 1974, 3, 135. 3. Wellings, F. M., and others. Lancet, 1977, i, 199. 4 Brown, T. J., Cursons, R. T. M. Scand. J. infect. Dis. (in the press). 5 Willaert, E. G. Ann. Soc. belge. med trop. 1974, 54, 429. 6 Anderson, K., Jamieson, A. Pathology, 1972, 4, 273. 7. Cursons, R. T. M., Brown, T. J. N.Z. med. J. 1976, 84, 479. 8 John, D. T., and others. Infect. Immun. 1977, 16, 817. 9 Chang, R. S., Owens, S. J. Immun. 1964, 92, 313. 10 Elridge, A. E., Tobin, J. O’H. Br. med. J. 1976, 1, 299. 11 Tew, J, and others. J. immun. Meth. 1977, 14, 231. 12. Visvesvara, G. S., and others. Am. J. trop. Med. Hyg. 1975, 24, 784. 13 Cerva, L, Kramar, J. Folia Parasitol. 1973, 20, 113.