1. The serum proteins of calves from birth to weaning, and the maternal colostral whey have been examined with the ‘classical’ Tiselius electrophoresis apparatus. Differences were shown between calves fed colostrum and those partially deprived of colostrum.

2. A study of the pre-colostral calf serum showed the presence of albumin and of two major components with mobilities similar to the α and β globulins of adult serum. A component forming approximately 1·4 % of the total serum proteins and with a mobility similar to that of γ1 or fibrinogen represented the γ1 globulin. This globulin component was not unconverted fibrinogen and may be autogenous γ globulin or γ globulin passively acquired in utero.

3. Autogenous γ globulin was evident in colostrum-fed and colostrum-deprived calves shortly after birth. The γ1 and γ2 components could be distinguished by the 10th day after birth, at which time the γ1 globulin was the greater. By the 30th day the γ2 globulin exceeded the γ1 globulin and a smaller component termed the γ3 globulin could usually be detected between the γ2 and the salt boundary.

4. Albumin concentrations generally fluctuated inversely to changes in the total serum globulins.

5. The α globulins associated with fetuin declined shortly after birth in the colostrum-fed group. In the deprived group α globulin first rose and then fell. In both groups minimum α globulin values were reached at about the 30th day, when the α1 globulin, although initially the major component in pre-colostral calf serum, was more depleted than the α2.

6. The β globulin frequently showed a transient though marked increase when the α globulins were at their lowest values.

7. No changes in the electrophoretic mobilities of the major serum proteins were detected as the calves matured, and no significant difference was found between the mobilities of the electrophoretic components of calf and adult sera.

8. The electrophoretic examination of colostral whey, colostral lacto-globulin fractions and calf serum immediately after suckling usually showed one lacto-globulin component. The relationship between the serum γ globulins and the lacto-globulin. is discussed.

The author wishes to thank Dr M. Robertson, F.R.S., and Sir Alan Drury, F.R.S. for their interest and encouragement during the course of this work, Dr W. R. Kerr for his co-operation in supplying most of the serum and colostrum samples, and Dr A. W. Stableforth and Dr J. S. Paterson for making certain cattle available for these experiments.

1. The lacto-globulin in bovine colostral whey, and the γ globulin in the serum contain the normal and induced agglutinins to the protozoon Trichomonas foetus.

2. A decline in γ globulin in sera from cattle approaching calving may be associated with the concentration of the lacto-globulin in the udder.

3. The elimination of the passive normal and induced agglutinins acquired by the neonatal calf was logarithmic. The autogenous production of γ globulin by the calf occurred soon after birth, resulting in an apparent decline in the rate of elimination of the lacto-globulins.

4. The autogenous production of γ globulin was not necessarily accompanied by that of detectable normal agglutinin.

5. Calves injected with T. foetus antigen within the first 3 weeks of life, and in which there was evidence of autogenous γ globulin, developed normal autogenous but not induced agglutinins.

6. Calves which received colostrum containing maternal induced and normal agglutinins, calves partially deprived of colostrum, and calves immunized in early life, produced the normal autogenous agglutinin between the 30th and 60th days, reaching the normal adult titre of 1/48–1/96.

7. The properties of the normal agglutinins were investigated. They possess some characteristics significantly different from those of the induced agglutinins. These suggested that the normal agglutinins were γ globulin molecules with an ‘accidental’ genetic configuration complementary to T. foetus.

The author wishes to thank Dr M. Robertson, F.R.S., and Sir Alan Drury, F.R.S. for their interest and encouragement during the course of this work, and Dr R. C. Campbell for the statistical analysis.

Two new statistical methods—one parametric, the other non-parametric—are developed to compare the slopes of quantal infectivity curves.

Ten influenza strains were examined with the aid of these tests, and the results show that virus yielded by the von Magnus phenomenon (i.e. containing ‘incomplete’ particles) and standard infective virus do not differ in the slope of their infectivity curves.

It is concluded that the von Magnus phenomenon yields a mixture of infective and non-infective particles; on this evidence Fulton's suggestion of uniformly changed virulence is untenable.

I am greatly indebted to Prof. P. A. P. Moran for helpful discussions of theory, and frequent corrections of statistical technique.

1. Coliform and paracolon β strains agglutinated to a high titre with OB antisera from two strains of Escherichia coli O111B4H2.

2. The presence of β antigen in some strains of Esch. coli O111B4H2 has been shown and therefore such strains should be classified as plus β forms. This factor could not be demonstrated by agglutination tests but was evident by its ability to produce and absorb β antibodies.

3. The dissociation and complex nature of β forms was discussed.

The writer wishes to express her thanks to Prof. S. D. Rubbo for his helpful criticism of the manuscript.

The effect of two chelating agents, sodium diethyldithiocarbamate and 8-hydroxy-quinoline (oxine), on 70 strains of brucellae (24 Br. abortus, 22 Br. melitensis and 28 Br. suis) has been examined.

The pattern of growth and inhibition of the three species as it appears in the carbamate test described by Renoux is of some value in the differentiation of species. Br. suis is more resistant to carbamate than the other two species.

The three species also differ in their sensitivity to oxine, and strains of Br. suis from various sources were consistently more resistant than the other species. This observation may be of use in the identification of Br. suis.

My thanks are due to Prof. Adrian Albert and Dr J. H. Marshall for helpful discussion of the chemical aspects of this work, and to Mr B. Madge, A.I.M.L.T., for technical assistance.

1. In an endemic area of Q fever in south-east England, two flocks of sheep, with serological evidence of infection by Rickettsia burneti, were examined for excretion of the rickettsia at parturition.

2. R. burneti was isolated from two out of ninety-six placentas collected from 118 ewes which lambed in one flock. The organism was also isolated from ‘wool tags’ obtained from one of the ewes which voided an infected placenta.

3. Infectivity titrations showed that the two placentas contained respectively 104·5 and 101·5 guinea-pig infective doses per gram of cotyledon. The wool tag from the sheep which voided the more highly infected placenta contained 103·25 infective doses per gram.

4. Except for one doubtful result, which could not be confirmed, no isolations were made from 194 placentas collected from 199 ewes in the other flock.

Details of the design of the microblender were supplied by Dr A. P. Goffe and Mr T. Nash of the Central Public Health Laboratory, Colindale, to whom we are much indebted. A gift of Triton X-100 was kindly given to us by the Rohm and Haas Corporation, Philadelphia 5 Pa.

1. Rickettsia burneti has been isolated from the tick Haemaphysalis punctata collected from sheep in an endemic area of Q fever in south-east England.

2. Isolation from one batch may have been due to contamination from the wool of sheep excreting R. burneti, but evidence is presented that isolation from another batch was due to the presence of the rickettsia in the tick tissues.

We should like to express our gratitude to Mr R. D. Brown, M.R.C.V.S., Mr F. L. J. Kett, F.R.E.S., and Mr J. Carey, who helped to collect the specimens. We are also indebted to Dr A. D. Lees (Agricultural Research Council Unit of Insect Physiology) for his help with the identification of the ticks. Finally we wish to acknowledge the valuable laboratory assistance given by Mr P. C. Collings, Miss Z. Page and Miss K. Arnold.

1. Particles of antu of 5 μ diameter were significantly less toxic to domestic rats than particles of 50–55 μ.

2. In each of three bio-assays they were less toxic than particles of antu of 100–110 μ.

3. It is concluded that the specification for antu for use in poison baits against R. norvegicus should not require it to be in the form of fine particles.

This work was initiated at the suggestion of Dr E. E. Turtle of this Division, and his advice during its progress is gratefully acknowledged. Thanks are also due to Mr L. E. Hammond for much help in the preparation of the test doses, and to Dr H. Egan and Mr E. I. Johnson of the Government Chemist's staff who carried out a number of the many micro-weighings involved.

The addition of glucose considerably reduces the amount of glycine required to protect complement against heat-inactivation.

A similar effect has been demonstrated with other sugars.

This paper describes efforts to determine the LD50 for mice, 24–48 hr. old, of pools of four strains of Herpes simplex virus, using the intraperitoneal route of inoculation.

The results have been analysed in an attempt to account for the very irregular relationship between dose of virus and mortality which was observed.

It is clear that, in addition to those difficulties which beset the interpretation of the results of any biological assay, titrations in very young animals raise difficulties of their own. The principal sources of difficulty are the high and uncertain non-specific mortality, the tendency of the mothers to eat their babies and individual variation between the litters. The latter, it is suggested, is mainly a reflexion of differences between the mothers, at least when a closely inbred strain of test animals is used.

Interesting and unexpected correlations were observed between the dose of virus and the proportion of mice eaten by the mothers and between the strain of virus and the proportion of mice eaten by the mothers.

It is emphasized that the results of titrations in very young animals must be interpreted with the greatest care and that in carrying out titrations the litter rather than the individual baby mouse should be regarded as the unit. Mice eaten by the mothers are probably best ignored altogether, although there are grave theoretical objections to doing this.

I should like to thank Professor H. R. Dean for his advice and encouragement. It is a pleasure to acknowledge the technical help which I have received throughout this investigation from Miss D. Sutton. I am also indebted to Dr M. G. P. Stoker for advice and for making available facilities in his laboratory; to Dr W. L. Smith and Mr T. H. Hollingsworth of the Department of Human Ecology, Cambridge, for advice on some of the statistical problems raised; to Miss M. E. Skinner and her staff on the ‘Animal Floor’ of this department, without whose help it would not have been possible to ensure that the large number of litters used were available when required; and to those of my colleagues who have been good enough to offer advice and criticism, especially Mr Douglas Annear.

1. In gastro-enteritis of babies caused by pathogenic varieties of Bact. coli the infection extends up the whole length of the small gut. Quantitative bacteriological examinations showed the contents of the upper intestine to harbour Bact. coli in numbers usually found only in faeces.

2. Bact. coli O 111, O 55, O 26, etc. are pathogens.

3. The syndrome of gastro-enteritis of babies associated with Bact. coli resembles cholera clinically, pathologically and bacteriologically.

4. There is a place in bacteriology for quantitative tests as distinct from qualitative tests.

I owe my thanks to many people: to Dr A. H. James for teaching me his methods of getting samples of duodenal juice; to Prof. A. G. Watkins, Dr P. T. Bray, Dr Emrys Harries, Dr S. Rosehill, Dr Bowen Owen and Dr W. R. Nash for allowing me access to their patients; to Dr T. E. Parry, Dr W. Mackay and Dr R. D. Gray for allowing me to attend and collect specimens from autopsies; it would be ungracious not to thank several ward sisters to whom I must have been a great nuisance. To my technician John Smith I am exceedingly indebted.

Enteric fevers are common in the Gold Coast, where their incidence is increasing owing to insanitary urbanization. Children and migrant labourers form the bulk of the patients. Typhoid fever accounts for three-quarters of the cases, but paratyphoid A and paratyphoid C also occur. The Widal test has attained renewed importance in early diagnosis since the introduction of chloramphenicol, and gave a presumptive diagnosis within 24 hr. in two-thirds of the cases. ‘Normal’ agglutinin levels were determined in 200 sera sent for the Kahn test, and provide data for interpreting the Widal test in the Gold Coast. Low titre H agglutinins in the sera of normal people are thought to indicate past vaccination or infection, but low titre O agglutinins are probably unrelated to either. Agglutination of a Salm. typhi H suspension by normal sera suggests that one adult African in eight in the Gold Coast has suffered from typhoid fever.

I have to thank Lt. Col. H. J. Bensted, O.B.E., M.C., for kindly providing Standard agglutinating sera.

My thanks are also due to Messrs D. Gellatly, A.I.M.L.T., K. C. Lisle, E. S. Martinson and A. T. Nartey for their technical assistance at various times during the past 2½ years.

I am indebted to Dr George Watt, M.B.E., acting Chief Medical Officer, Gold Coast, for permission to publish this paper.

Lanoline-water cream containing 1% of antibiotic was applied for 7–14 days to the anterior nares of 124 carriers of antibiotic sensitive staphylococci. Nose swab cultures, examined during the administration of the cream, were negative for Staph. pyogenes in all carriers who received oxytetracycline, and in 80, 75, 70 and 67% of those who received penicillin, chlortetracycline, streptomycin and chloramphenicol respectively.

Seventy per cent of carriers gave nose-swab cultures negative for at least 2 weeks, 50% for at least 4 weeks and 25% for at least 20 weeks after the cessation of treatment. Over the same period only 1% of a control group ceased to be carriers.

The nares of 20% of carriers receiving antibiotic were temporarily colonized with antibiotic-resistant strains, but in only one case (0·8%) did the resistant organism persist throughout the period of examination.

The use of this method of controlling staphylococcal carriage in preventing staphylococcal infection is discussed.

I wish to thank Prof. T. J. Mackie for his advice.