1. The sizes of the droplets and droplet-nuclei produced by sneezing, by coughing and by speaking, were studied by the microscopic measurement of 12,000 droplet stain-marks found on slides exposed directly to mouth-spray, and of 21,000 stain-containing droplet-nuclei recovered from the air on to oiled slides exposed in the slit sampler.2. From these measurements it was calculated that the original diameters of the respiratory droplets ranged from 1 to 2000 μ, that 95 % were between 2 and 100 μ and that the most common were between 4 and 8 μ. Similar size distributions were exhibited by the droplets produced in sneezing, in coughing and in speaking, except that, in the case of sneezing, the smaller droplets were relatively more numerous.3. The respiratory droplet-nuclei were found to range in diameter from ¼ to 42 μ; 97 % had diameters between ½ and 12μ; the commonest diameter was between 1 and 2 μ.4. The proportion of droplets of each size which will contain bacteria, whether commensal or pathogenic, is determined by the size of the droplets and by the numbers of bacteria in the secretions atomized. Calculations made on the basis of the size distributions obtained in this investigation indicated that few of the smaller droplets, and thus few of the droplet-nuclei, are likely to contain pathogenic organisms. Droplet-spray is unlikely to give rise directly to true airborne infection unless very large numbers of pathogenic organisms are present in the secretions of the anterior mouth.5. The persistence of droplet-nuclei in the air of a 1700 cu.ft. room and of a 70 cu.ft. chamber was investigated by sampling the air with the slit sampler at intervals following sneezing.6. When the air was not artificially disturbed by a fan, the time taken for the disappearance from the air of 90% of the bacteria-carrying droplet-nuclei varied from 30 to 60 min.; the nuclei larger than 8 μ in diameter usually disappeared within 20 min., and the nuclei larger than 4 μ within 90 min.; the smaller nuclei, few of which contained bacteria, remained airborne for much longer periods, on one occasion for at least 30 hr. When a fan was run throughout the experiment, the nuclei disappeared from the air much more rapidly.

No abstract

Berry, L. Joe (Bryn Mawr College, Bryn Mawr, Pa.), Dorothy S. Smythe, and Louise S. Colwell. Inhibition of inducible liver enzymes by endotoxin and actinomycin D. J. Bacteriol. 92:107-115. 1966.-Bacterial endotoxin at the ld(50) level lowers liver tryptophan pyrrolase in mice, it prevents for 16 to 20 hr the induction of the enzyme by a concurrent injection of cortisone, it lowers significantly but does not prevent substrate induction, and it reduces the enzymatic activity promptly and significantly when administered during the course of hormonal induction. The ld(50) amount of actinomycin D has a similar effect on tryptophan pyrrolase, except that its inhibition of induction by cortisone persists for a longer period of time. Endotoxin in the intact mouse induces tyrosine-alpha-ketoglutarate transaminase almost as well as cortisone, but not in the adrenalectomized animal, a fact that suggests induction of this enzyme is due to release of endogenous adrenal hormones. Actinomycin D, on the other hand, has an effect on this transaminase similar to that on tryptophan pyrrolase. The site of action of endotoxin and actinomycin D would appear to be similar for one of the two enzymes studied and different for the other, a relationship that requires a specificity difficult to imagine for a material as complex as endotoxin.

No abstract

PLATES LXVIII-LXX)NON-FLAGELLAR filamentous appendages about 0.01 p in width were demonstrated electron-microscopically in various Gram-negative bacilli by Houwiiik and van Iterson (1950). These filaments were named " fimbriz " by Duguid et al. (1955), who showed that in Bacterium coli their possession conferred on strains the power of adhering t o and agglutinating red blood cells. A similar association of hzmagglutinating activity with fimbriation was found in Bacterium c l o a m by Constable (1 956). The electron-microscopic demonstration of fimbriz in replica preparations of living Bact. coli on agar (Schreil and Schleich, 1955) removes earlier doubts that they might be artefacts produced on drying from amorphous slime. J. PLTH. BACT.--YOL LXXIV (1957 397 39s J . P. LIUGUID AND R. 22. (1954) found that fimbriation reduced the susceptibility of a Bact. coli strain t o certain T bacteriophages ; these phages were inactivated by a preparation of detached fimbriae.BILLIESThe occurrence of hzmagglutinating activity in many pathogenic bacteria suggests that the pertinent adhesive properties may have their natural function in fixing the bacteria or tjheir toxins to the surfaces of host cells. Keogh and North (1948) obtained evidence that the virulence of H . perfussis strains was related to their content of haemagglutinin. However, it was later shown that, antibodies for this h~emagglutinin are not protective (Masry), the main protective .antigen being another substance which adsorbs on red cells without agglutinating (Pillemer, 1950).Except for the exotoxin-producing species, lit,tle is known of the mechanisms whereby pathogenic bacteria initiate their attack on the epithelial cells of the respiratory and alimentary mucosze. I n our present study of the occurrence of fimbria and hzemagglutinins in Shigelka, we sought evidence that these might fulfil some role in the pathogenesis of dysentery. MATERIALS AND METHODS Strains.We examined 247 strains of various Xhigella species (tables I and 11). The 124 strains designated '& recently isolated ", were examined within a few subcultures (usually 4-7) and a few weeks or months after isolation; 65 were isolated in Edinburgh during 1955 and 1956, while the remainder originated from many countries and were mostly obtained from Dr K. P. Carpenter, Dysentery Reference Laboratory, London. The 123 " older straim " were ,obtained from two main sources. Dr J. S. K. Boyd supplied 34 strains isolated between 1936 and 1951, and maintained as slope cultures. A further 86 strains were obtained from the National Collection of Type Cultures ; most of t,hese had been isolated many years ago; 21 received before 1935 were originally subcultured 6-12 monthly, but later freeze-dried ; 25 received between 1936 and 1949, and 40 received more recently, were freeze-dried on receipt. Among 154 strains of known origin, 100 were from 8 European countries, 21 from India, 15 from Africa, 2 from Arabia, 7 from the Far East and 9 from the Americas. All strains were identified by biochemical and sero...

No abstract

THE epidemiological value of different methods of distinguishing types within a bacterial species depends on their discriminating ability, reliability and ease of performance. Discriminating ability depends primarily on the number of types distinguished, but is poor if a large proportion of strains falls into a single type. Reliability depends on the reproducibility of each isolate's results in repeated tests and on the stability of the typing characters among different isolates from the same epidemic episode.Probably the best single method for differentiating types in Salmonella typhimurium is the phage-typing system of Callow (1943, 1951) , 1972). This method is less discriminating and more laborious than phage-typing, but it distinguishes different biotypes within individual phage-types, so that its combination with phage-typing gives a finer discrimination than either method alone (Kallings and Laurell, 1957;Rische and Kretzschmar, 1962;Lewis and Stocker, 1971).In a combined phage-typing and biotyping study of S. typhimurium we found some serious defects in the Kristensen scheme.(1) The fermentation tests with d-, Z-and rn-tartaric acids and citric acid proved unreliable and often gave different results on different occasions of testing the same isolate. Improved, more reliable methods were developed for the tartaric acids by Alfredsson et al.(1972) and these methods are used in the new scheme described in the present paper.(2) In our series of strains, both the original and the modified tests identified several biotypes with patterns of reactions different from those of the twenty-one types already recognised, but the new types could not be accommodated in the Kristensen classification in such a way as to show their relations to the existing types. For example, a group of strains designated " type 19Xd " by Morgenroth and Duguid (1968) resembled biotype 6 except that it was rhamnose negative; if this new type had been added to the scheme as type 22, its close relation to typ6 6 would not have been apparent. (3) We found that discrimination in biotyping could be improved by the addition of tests for the fermentation of trehalose in peptone water, fermentation of rhamnose in Bitter's medium, fermentation of inositol at 25"C, gas production from glucose, growthfactor requirements and presence of flagella and fimbriae, but the results of these extra tests could not be shown in any convenient way within the Kristensen classification. Mutational fermentation. Alfredsson et al. (1972)found that a major cause of unreliability in tests with the tartaric acids was the power of many non-fermenting strains to produce ~ ~~

SUMMARY: F'imbriae were found in 125 of 154 non-motile Klebsiella strains examined by electron-microscope in serial aerobic broth cultures. Fimbriate strains occurred in each of the capsule serotypes 1-72 and mostly showed the biochemical reactions of saprophytic KZebsieZZa aerogenes. The fimbriae were clearly distinguishable from the capsules and occurred also in non-capsulate mutants. Most fimbriate strains showed evidence of varying reversibly between a fimbriate and a non-fimbriate phase. The proportion of bacilli with fimbriae was greatest in broth cultures and was decreased, though never eliminated, by serial cultivation on nutrient agar. All 29 permanently non-fmbriate strains belonged to serotypes 1-6 and showed biochemical reactions common in pathogenic K. pneumoniae, K . oxaenae and K . rhinoscZermatis strains (anaerogenic, methyl-red positive, non-citrate utilizing or nonlactose fermenting). They differed from the fimbriate strains ingrowing less abundantly and usually in failing to form a pellicle on broth.All the fimbriate strains, but none of the non-fimbriate, showed one or other or both of two kinds of adhesive property : one attributed to an 'MS adhesin ', susceptible to inhibition by D-mannose and associated with a thick variety of fimbriae; the other due to an 'MR adhesin', resistant to mannose and associated with thinner fimbriae. Bacilli with MS adhesin rapidly adhered to red blood cells of the guinea-pig and other animals, except the ox, t o leucocytes and to intestinal epithelial cells, including those of ox, to Candida albicans cells, to the mycelium of Aspergillus niger and other moulds, and to plant root-hairs. Bacilli with only the M R adhesin did not adhere to untreated red cells, leucocytes or 'smooth' yeasts, but adhered rapidly to ox and other red cells treated with tannic acid, red-cell stromata heated a t 70' or loOo, fungus mycelium and plant mot-hairs. B a c i l l i of many fimbriate strains and one non-capsulate non-fimbriate strain adhered to glass and cellulose.