In previous papers (1922, 1924) we have investigated the capacity possessed by certain micro-organisms of reducing sulphites to sulphides, and have applied it (1) to the differentiation of bacteria, and (2) to the bacterioscopic analysis of water.

The blood picture in rubella is characterised by the presence of numerous plasma and Türk cells and by the absence of toxic changes in the neutrophilic leucocytes.

Plasma cells occur also in scarlatina and in measles but with less regularity and, as a rule, in smaller numbers, but these diseases are always associated with “toxic” changes in the neutrophilic cells.

Such haematological features must be considered in relation with the clinical findings because plasma cells and toxic changes in the neutrophiles may occur in entirely different conditions, e.g. in association with the lympho cytosis following almost any acute infective malady.

The haematological method would probably find its greatest application in those cases in which the characters of the rash are not sufficiently well developed to enable a diagnosis to be made in the usual way.

Acknowledgments. My investigations were carried out at the Institute of Pathology of Charing Cross Hospital Medical School, and I am much indebted to the Director, Dr Piney, for assistance and advice, and to Dr Whittaker of the Metropolitan Asylums Board for access to clinical material. Moreover, financial assistance has been afforded by The Virol Research Scholarship which I hold.

Though we have, in former papers (1915, 1916, 1920), given details of the apparatus we used in producing sterilisation of milk by the electric current, it will be a convenience to readers to repeat some facts in regard to the arrangements of the electrodes and electrical circuit. The quantity of current used depends on the size of the apparatus, arid the voltage is normally between 3000 and 4000. The alternating current is carried by the milk and is applied by means of three copper electrodes, ⅛th of an inch thick, each enclosed in a glass electrode chamber.

1. A short account is given of the occurrence of plague epidemics in wild mammals other than “domestic” rats and mice. Enough is now known to show that such epidemics are often of great practical importance.

2. The method by which most rodents regulate their numbers is as follows: increase in numbers over several years up to a point at which an epidemic of some sort occurs, which kills off a large proportion of the population. Increase then takes place again, and is followed by another epidemic, and so on indefinitely.

3. These periodic fluctuations are probably controlled by widespread climatic fluctuations, the best evidence for this being that in certain cases the former run synchronously in widely separated countries.

4. There appears to be a dominant short period in fluctuation of three to four years, and a larger movement of period ten to eleven years, both in North America and Europe. The fact that the 11-year sunspot cycle roughly corresponds with the larger movement is significant.

5. The data available from Eastern Asia suggest that there too, small mammals fluctuate periodically in numbers, and with similar periods to those of North America and Europe. There is also some evidence that the maxima of the 10–11-year fluctuations coincide in Eastern Asia and in North America, just as those of the 3–4-year period coincide in Northern Canada, Greenland and Norway.

6. Evidence is given that the plague marmot (Arctomys bobac), and other rodents carrying plague, are liable to these fluctuations. If this proves to be true, it may be possible, when fuller data are available, to forecast with some accuracy the years of epidemics among these animals, and if this can be done we shall have some means of gauging the likelihood of the occurrence of plague outbreaks in the human population of those regions.

7. The available data are admittedly fragmentary, but it is probable that between the years 1931 and 1934 epidemics among A. bobac in Transbaikalia and Mongolia will be severe, and that these events will lead to an increase of plague mortality in man.

Enquiries have frequently been received at the National Institute for Research in Dairying concerning the efficiency of the Fermentation-Reductase Test as a method for the grading of milk, and the suggestion has been made that this test, which is now in use in some European countries, might with advantage be substituted for the standard methods of counting bacteria and demonstrating the presence or absence of B. coli, which at present form the basis for the grading of milk in England. A study of this question was therefore undertaken, and since similar enquiries are still being received, it is thought advisable to publish a résumé of the results that have already been obtained.

The question of non-specific formation of antibodies is of considerable theoretical and practical importance, and is closely related to the problem of “non-specific immunisation” which has attracted attention in therapeutics during recent years.

Any conclusions that are reached in such an enquiry as this can only be put forward very tentatively, for they are based (1) upon limited data which cannot be free entirely from error, and (2) upon the averaging of individual results drawn, it is true, from the same class, but from families varying considerably in economic status. For this reason, without much further investigation, any generalisation would be out of place, and the conclusions summarised below are put forward not as definite results but only as indications of the actual facts.

(1) In view of the fact that minimum wage legislation for agricultural workers is being introduced in answer to the statements that the economic status of these workers is deplorably low, it is important to bring forward any facts relating to the actual condition of the rural families.

(2) For an investigation of these conditions, Essex (rural districts) was chosen as the field of enquiry in 1923, and for purposes of comparison and confirmation of results earlier rural investigations have been utilised.

(3) A study of the experimental work on metabolism suggests that the agricultural labourer would certainly need a gross amount of at least 3300 calories per day, and that this assumption will not be likely to err on the side of excess.

(4) The Essex studies of 1923 give a mean value of 2872 calories per man value per day and only 17 per cent. of the families secure over 3300 calories per man per day.

(5) This result is not at variance with the conclusions reached in earlier investigations in rural areas, and the suggestion accordingly is that underfeeding is, to some extent, present in the rural areas. (Lusk writes “it is evident… that one may reduce the basal requirement of energy by undernutrition, and that this process may largely economise food, but that to accomplish a given amount of work a fixed amount of food-fuel is required, irrespective of the nutritive condition of the organism.”)

(6) This under-feeding occurs although nearly three-quarters of the total income is spent on food (leaving a very small amount for other essentials and practically nothing for non-essentials), and although the diet is substantially increased by the consumption of home produce. Without the latter the diet would be seriously deficient.

(7) In the Essex studies nearly one-third of the money spent on food was devoted to bread and flour.

(8) On the other hand, in comparison with urban studies, the rural diet compares very favourably with that of urban workers possessed of a higher money income.

(9) This advantage to the rural family is gained by (a) a larger percentage expenditure on food; (b) a less varied diet containing a substantially smaller amount of meats and hence less animal protein; a substantially larger amount of breadstuffs and hence more carbohydrates in place of fats1; (c) by the consumption of home produce, which, however, entails, in its turn, a lengthening of the day's work and increased needs for food. Without this home produce the diets would compare very unfavourably with those of the urban workers studied. With it, the Vitamine content, as well as the energy value of the diet, must be substantially increased.

(10) In spite of the diet being low there is little evidence of malnutrition amoung the rural children of Essex, and their physique is better, age for age, than that of urban children, living on a diet equivalent or superior in energy value. Other factors, especially the “racial” factor and the “fresh air and sunlight” factor, however, must exert such an influence that the efficacy of the diets cannot be satisfactorily measured by a comparison of heights and weights.