The present communication deals with an analysis and comparison of the principal physical measurements, physical characters and relative bodily proportions in groups of boys and girls at corresponding ages who were under treatment for asthma and rheumatism respectively in the special clinics established at the Sick Children's Hospital, Great Ormond Street, and of other groups of boys and girls of similar ages with no history of asthma or rheumatism and no obvious serious impairment of general health who were attending the Outdoor Department of the hospital for treatment of trivial ailments. The last-mentioned series of boys and girls have been assumed to represent the normal hospital population and have been referred to throughout as the normal groups. The object of the investigation was to determine if there was any real evidence that the asthmatic and rheumatic children differed from one another or from the normal children in general physical type or build. The inquiry has been restricted to children in the age period 6–12 years, and these have been taken in two-yearly age groups in order to provide more adequate numbers of observations for comparison. In all, the characters of approximately 1212 children have been dealt with. Of these, 368, 236 boys and 132 girls; 459, 189 boys and 270 girls; and 385, 195 boys and 190 girls, were considered to be asthmatic, rheumatic and normal respectively. A study of the frequency distributions, mean values and variabilities of the characters compared at the corresponding ages in the three groups of boys and girls seems to provide a reasonable basis for the following conclusions:

1. The corresponding averages of practically all the absolute physical measurements brought under review are in such close agreement in the contrasted groups of asthmatic and rheumatic children that the data provide no evidence of any real difference in the general dimensions of children with the respective diseases. Though the average antero-posterior diameter of the chest is probably greater in the asthmatic than the rheumatic children, no emphasis can be laid on this feature as it may be, in part, the result of the disease.

2. Though at some ages the mean dimensions of a few characters in the asthmatic and rheumatic groups do differ from the corresponding averages in the normal series, the differences are not sufficiently consistent at the different ages and in the two sexes to indicate any real divergence from the normal.

3. A comparison of the relative proportions of the body as expressed by the percentage ratios or indices, stem length/stature, body length/stature, biacromial diameter/body length, bicristal diameter/biacromial diameter and length of arm/length of leg, at corresponding ages in the asthmatic and rheumatic children does not provide any definite evidence that children suffering from these two diseases differ from one another in morphological type.

4. Though occasionally, at one or more ages, one or more of the indices used to express the relative proportions of the body may appear to differ in mean value from the normal averages at the corresponding ages the differences are not sufficiently consistent in regard to age and sex to suggest any real divergence from the normal.

5. Comparison of the frequency distributions of hair colour shows that the asthmatic children do not really differ from the rheumatic children in respect of pigmentation. The rheumatic children do not differ from the normal, but both the asthmatic boys and girls show a relatively greater excess of the blonde haired type and a greater deficiency in the dark haired as compared with the normal than might be expected to occur as a chance variation.

6. With regard to distribution of eye colour, the asthmatic boys do not differ from the rheumatic boys, but the asthmatic girls appear to have a relative defect in brown eyes and a relative excess in blue eyes as compared with the rheumatic group; neither in boys nor girls with asthma or rheumatism can the distribution of eye colour be said to differ significantly from the normal on the data available.

7. There is no definite evidence that the proportions of relatively long and slender and short and thick-set types, as appraised by general inspection, amongst asthmatics differ from the normal.

8. There is no evidence in the data for boys or girls that the distribution of grades of posture differs significantly in either the asthmatic or rheumatic groups from the normal, though in girls with rheumatism the habitual posture appears to be relatively worse than in those with asthma.

9. Both in regard to form of face, as expressed by the upper and total facial indices, and form of head, as represented by the cephalic index, the groups of asthmatic and rheumatic children do not differ significantly from one another or from the normal.

10. The differences in the three groups of children, the asthmatic, the rheumatic and the normal, in respect of the aggregate of physical characters brought under review are relatively so slight that they cannot be considered to support the view that asthmatic and rheumatic children really differ on the average from one another or from the general population of children from which they are drawn in bodily conformation or physical type, though they may possibly, indeed probably, differ in other constitutional traits.

1. A photo-electric cell was used to measure photometrically the opacity of cultures of Salmonella gallinarum growing in bouillon.

2. The opacity of a culture was defined as the number of 24-hour-old bacteria which when suspended in the same type of bouillon as those under investigation would absorb the same amount of light. Seven experiments were made to compare the opacities of growing cultures with their actual numbers throughout the first 24 hours of growth. An additional experiment was made to find directly the numbers of bacteria of different ages which would absorb the same amount of light.

3. The ratio between opacity and number was greatest at about 3 hours, decreasing to unity at about 7 hours; the optical behaviour of cultures 7 hours old was thus the same as of cultures 24 hours old. It is possible from the results that the ratio decreased to a minimum at about 9 hours, thereafter again rising to unity. The value of the greatest ratio was greater as the inoculum was less. The mean greatest ratio was 4·6, the mean time being 3 hours. This means that it would take about 4·6 times as many 24-hour-old bacteria as 3-hour bacteria to absorb the same amount of light.

4. It was shown that the variation in the opacity-number ratio was in all likelihood due to changes in the size of the organism during growth.

5. The results suggest that the opacity curves give a truer picture of actual growth than the ordinary growth curves, since the cells change in size during growth.

1. 1000 urinary and 500 faecal cultures of coliform bacteria were studied by means of the methyl-red, citrate and indol tests and also by their fermentation reactions upon saccharose and dulcite.

2. 8 per cent, of the faecal strains and 52 per cent, of the urinary strains were of the aerogenes type, i.e. methyl-red negative and citrate positive. Indol was produced in 94 per cent, and 43 per cent, of cases respectively.

3. This would suggest additional evidence that the coli group are quickly outnumbered by the aerogenes group once they leave the faeces, and also that the presence of the aerogenes type is not confined even in large numbers to an extra-corporeal habitat.

4. In water bacteriology the constancy or change in the relative proportions of coli and aerogenes strains present is the goal at which to aim. The aerogenes variety cannot of itself be regarded as non-excretal or even non-faecal.

5. From a sanitary standpoint, the classification of coliform bacteria by means of the methyl-red and citrate tests cannot be regarded as entirely free from error so long as (a) the aerogenes strains are found in faeces, in however small numbers, (b) these strains occur as the predominant type in urine, and (c) intermediate strains are found in the faeces, urine, soil and water.

6. The principal value of these tests lies in their comparative specificity in identifying B. coli of immediate faecal origin. The presence of such an organism, interpreted together with the indol test, may be said to suggest dangerous faecal pollution if found prominently in a water sample. The presence of other types identified by the same tests cannot be regarded as having the same significant value in regard to a negative opinion.

7. A plea is made for a further study of coliform bacteria by means of the tests described in relation to the biological behaviour of the bacteria under different environmental conditions.

1. A systematic study has been made of 165 Brucella strains from different parts of the world, and a less intensive study, comprising only the H2S formation, dye sensitivity, and agglutinin-absorption tests, of a special group of 156 strains from the north-east, east, and south-east of France.

2. For purposes of differentiation little weight can be attached to the use of morphological appearances, abundance of growth in culture, pigment formation, the appearance of crystals in the medium, the production of alkali in peptone water, or the utilisation of glucose.

3. The presence of 5–10 per cent, of carbon dioxide in the atmosphere is essential for the growth of most freshly isolated bovine abortus strains. This concentration frequently favours the growth of melitensis, though not of porcine abortus, strains. A concentration of 40 per cent. CO2 almost invariably inhibits to some extent the growth of melitensis and porcine abortus strains, and frequently inhibits the growth of bovine abortus strains. It may be concluded that, while the inhibition of growth of a given strain by 40 per cent. CO2 cannot be considered of differential value, the fact that growth in 40 per cent. CO2 is as good as, or better than, that in air is definitely in favour of its being of bovine abortus type, while the occurrence of growth in 10–40 per cent. CO2 but not in air almost certainly indicates that the strain belongs to this type.

4. Most bovine abortus, para-abortus, and American porcine abortus strains produce H2S freely for 3 or 4 days, while melitensis, paramelitensis, and Danish porcine abortus strains fail to produce any, or more than a small quantity on the first day only. Though the failure of a given strain to produce H2S cannot be regarded as indicating that it is of melitensis, paramelitensis, or Danish porcine abortus type, the definite production of H2S can be considered as very strong evidence that it is of bovine abortus or American porcine abortus type.

5. The dye sensitivity method, introduced by Huddleson, is of very real value in differentiation, and is the only certain method available for distinguishing between the bovine and porcine abortus types. By its means it is possible to divide Brucella strains into three main groups—bovine abortus and para-abortus, porcine abortus, and melitensis and paramelitensis. Not all strains, however, within a given group behave alike, and a number of sub-groups can be established on the basis of special sensitivity to one or more dyes. The method not infrequently yields results that demand considerable experience in their interpretation, and it cannot always be relied upon for the correct typing of individual strains.

6. The thermo-agglutination test is one of the simplest methods of detecting antigenic roughness. Any strain that fails to react negatively to this test is unsuitable for the production of smooth antiserum or for use in routine diagnostic agglutination work.

7. The agglutinin-absorption method, performed by a strictly quantitative technique, enables a differentiation to be made between bovine and porcine abortus strains on the one hand, and melitensis strains on the other, provided that smooth strains are employed both for the preparation of antisera and for absorption. The use of direct agglutination by monospecific abortus and melitensis sera affords a rapid and accurate means of typing individual strains, and may prove of value in the examination of strains which, while partly rough and unsuitable for absorption experiments, still retain sufficient smooth antigen to be agglutinated by one or other serum. It is also of great service in the detection of mixed strains. Both methods yield identical results.

8. In the examination of the main group of strains the serological method proved more valuable than any other method in the correct allocation of individual strains, and the results agreed closely with those afforded by the dye method. In the examination, however, of the special group of strains from the north-east, east, and south-east of France, there was frequently a marked disagreement between the results of the H2S and dye tests on the one hand, and the serological method on the other, the chief divergence being that a number of strains reacting by the former methods as melitensis behaved serologically like abortus. It appears as if, in this particular area of France, strains occur having the metabolic properties of melitensis and the antigenic constitution of abortus.

9. A study of the main group of strains by the various methods enumerated enabled them to be classified into the following groups: (a) bovine abortus with five sub-groups, (b) porcine abortus with two sub-groups, (c) melitensis with two sub-groups, (d) para-abortus with two sub-groups, and (e) paramelitensis.

10. A study of the special group of strains from the eastern districts of France revealed the presence of ten sub-groups. Since it is rather doubtful to what main group many of the strains belong, they have been classified according to the host from which they were isolated.

Since Anderson, Happold, McLeod and Thomson (1931) described their division of the varieties of Corynebacterium diphtheriae into three groups—gravis, intermediate and mitis—and submitted correlations with clinical types of diphtheria, several workers have published results of investigations which, for the most part, are against such a hard and fast classification.

In our investigation of our subject we have perused the literature, analysed the data of other workers, conducted a number of animal experiments, considered the bearings clinical and pathological of a case of presumed toluene poisoning, and conducted an investigation into the blood state and symptoms of co-workers with this case and yet are unable to give a categorical answer to the question. What are the relative toxicities of toluene and benzene? We would almost feel inclined to echo the words of another worker in the experimental field with regard to benzene, Neumann (1915), who concludes: “The fact that my results so often do not correspond either with one another or with those of Selling, so often indeed do not correspond to expectation, justifies the conclusion that, although benzene is to be regarded as a powerful leucotoxin, it manifests great individual differences in its mode of action and it is very difficult to lay down exactly what action is to be expected.”

We should be inclined to put the case somewhat differently however and put forward the following view as our tentative finding.

The actions of benzene and toluene are very similar. That of benzene is the more powerful because it is a more volatile substance and can therefore reach higher concentration than toluene. This greater volatility at the same time is a reason why it is eliminated more quickly and its effects are less lasting than those produced by toluene. The action of both these substances is on young body cells, probably whatever they are and wherever they may be found, but is most easily discoverable because of its action on the young bone marrow cell and the consequent effect on the cells of the peripheral blood. It is not a specific action but operates, according to circumstances, on the stem cells of the myeloid leucocyte or the erythrocyte. This does not mean that the substances have little or no effect on the lymphocyte, for the lymphoid tissue in the body is so very abundant that it would only be in extreme cases that the effect on that cell would be manifest peripherally. Like many poisons, indeed like many of the drugs which are used in Medicine, the first effects of these substances are stimulating. Leucocytosis, erythrocytosis and possibly thrombocytosis are early manifestations or the manifestation of slighter action, with histological evidence of hyperplasia. This is succeeded by the truly toxic or destructive action and the production of leucopenia, anaemia, thrombopenia and the effects due to action on other organs. The ultimate result in clinical terms is an agranulocytic anaemia accompanied by purpura haemorrhagica and, as the pathological manifestation an aplastic bone marrow. The fatal termination is commonly an infection to which there is no resistance possible in an individual deprived of defence cells and possibly also of defensive substances in the body fluids. Powerful solvents of fat as both these chemical substances are, they produce an action through this means also, on nerve cells, capillary endothelium and parenchyma of organs with consequences represented in nervous symptoms and in haemorrhages. This then is our reading of the results of our investigation and explanation of the variability which may be found, a variability which is regarded mainly as one of phase, although we do not exclude idiosyncrasy altogether as explanation.