1. Numerical counts on gelatin media are of very limited value for surface well examinations and are quite useless for open draw-wells. The blood-heat enumeration is of use, but still of but limited value.

2. The influence of rainfall and local conditions generally is marked for this class of waters with regard to their bacterial content.

Whether the well is uncovered, or covered and provided with a pump has a very important influence upon the bacteriological results. For a good many wells which show very bad bacteriological findings the cause of pollution is frequently due to surface pollution rather than to pollution of the ground water through the soil. If these wells are properly protected they will become suitable sources of drinking supply.

3. The results obtained confirm the value of B. coli and streptococci estimations as the best tests by which to judge the freedom from pollution of such waters.

The streptococci results are of great value, only second in importance to the B. coli determinations. In numerical distribution these two organisms closely agree. The presence of streptococci is more reliable as evidence to deduce pollution than their absence is to exclude it.

4. An opinion regarding the freedom of surface wells from pollution is best formed on the basis of combined topographical and bacteriological investigation.

5. Inspection alone is frequently quite insufficient for the purpose of determining if a well is contaminated and if it should be closed.

Of the 50 wells examined no less than 31 had to be classed as doubtful (Series C), topographical examination alone affording insufficient data for the expression of a decided opinion.

In view of the fact that pure water supplies are not easy to obtain in many rural districts, a water supply, which may have been in use for many years, cannot be condemned in the absence of clear evidence that it is unsatisfactory and dangerous.

6. In surface well waters a large proportion of the coli-like organisms isolated are atypical in one or more particulars.

7. Typical B. coli implanted into soil showed some alteration of character, but the changes were not extensive and no evidence was obtained that the widely aberrant organisms met with in different soils and waters ever represent typical B. coli altered by unfavourable environment.

8. Organisms closely allied to B. coli, but differing in one or more characters, possess significance as indicators of faecal contamination.

The more nearly the organism isolated resembles an “excretal” B. coli the greater is its significance as an indicator of pollution. Consequently the fewer the number required to condemn a sample water in which they occur. Stated as a working proposition, the more the characters of the coli-like organisms deviate from that which for convenience may be spoken of as the typical form, the greater the proportionate number of them required to condemn the water.

9. The presence of ‘excretal’ B. coli in 10 c.c. or less of a surface well water points to undesirable pollution and is sufficient to condemn the water.

1. It is inadmissible to study the “Danysz Effect” on tetanolysin owing to its rapid deterioration.

2. The so-called “equivalents” of toxin and antitoxin deduced by Arrhenius and Madsen are arbitrary.

3. No evidence has yet been advanced that the “Danysz Effect” has a limiting value when the time of contact of the first fraction of toxin with the antitoxin is prolonged.

4. If a limiting value of the “Danysz Effect” exist that calculated by Arrhenius is probably erroneous.

5. The monomolecular formula used by Arrhenius is merely an interpolation.

6. The “Danysz Effect” is much better represented by a bimolecular formula.

7. No confirmation of the “equivalents” of toxin and antitoxin has as yet been obtained from the “Effect.”

8. Expediency appears to be the only justification for assuming that the “Effect” is due to either a modified antitoxin or to a modified toxin, viz. epitoxonoid.

9. All the phenomena of the “Effect” hitherto advanced have their counterpart in the staining of paper, porcelain, etc., by anilin dyes.

1. The relation between lethal doses of diphtheria toxin and the times in which they kill guinea-pigs, or lethal times, is approximately such that the lethal dose multiplied by the corresponding lethal time gives a constant value—a hyperbolic relation.

2. Deductions with regard to lethal doses from deaths occurring on the first day or on any day after the sixth give such widely divergent results that, in our opinion, they are at present of negligible value.

3. Under the ordinary circumstances of standardisation of diphtheria toxin and antitoxin the relation given in conclusion No. 1 is as close an approximation as any alternative likely to prove of practical utility.

4. The graphic representation of L+ doses against lethal times gives a straight line relationship.

5. Five of the best investigated toxins confirmed Ehrlich's views in so far as that with toxicities diminishing by 50% their neutralising powers remained practically unimpaired.

6. The individual sensitivities of guinea-pigs to free diphtheria toxin render any general relation between lethal times and doses of little value when a small number, e.g. 5, of test animals is inoculated.

7. Even with a dose of toxin which would, in any ordinary series of tests, be regarded as an L+ dose, 5 out of 33 guinea-pigs survived the sixth day or 15%1, and the m.l.d. of a well-investigated toxin showed similar variations; consequently both the L+ and m.l.d. values of a toxin are not those amounts which kill with certainty in a fixed time, but those which will cause death in that time with the greatest probability.

8. After inoculation, guinea-pigs unless suffering from the consequent “shock,” continue to increase in weight throughout the first day, even with doses that may subsequently prove lethal, and with highly toxic doses the variation in weight during this time is of little significance.

9. The greater the amount of free toxin injected, the more rapid is the increment in weight followed by a decrement.

10. The weight of a test animal at the end of 24 hours after inoculation forms a better normal or origin with which to compare the subsequent time-weight relations than the initial weight before injection.

11. On the basis of Conclusion 10 the general connection is found that with increasing amounts of free toxin the position of the maximum loss of weight is gradually transferred from the second to the fifth day for guinea-pigs which survive over six days.

12. The time-weight ratios for almost certainly lethal doses give straight line connections, corresponding probably to starvation curves of the guinea-pig.

1. Nickel carbonyl is a highly toxic compound.

2. Dissociation of nickel carbonyl takes place rapidly at the temperature of the body in the presence of air, moisture and carbon dioxide.

Nickel carbonyl also dissociates at the temperature of the body in the absence of air, moisture and carbon dioxide, but not to the same extent as in the presence of these owing to the fact that a reversible action occurs between the nickel and the dissociated carbon monoxide.

3. Nickel carbonyl is soluble in water to the amount of 2·36 c.cm. of vapour (i.e. 6·43 mgrs. of nickel) per 100 grs. at 9·8° C.

The amount dissolved in water is directly proportional to the pressure, when the temperature is constant.

The solubility in water diminishes as the temperature is raised.

4. The solubility of nickel carbonyl in serum and in blood is about 2½ times greater than that in water. At 10° C. 100 grs. of serum dissolves 5·98 c.cm. of nickel carbonyl vapour (i.e. 15·71 mgrs. of nickel).

5. When nickel carbonyl is brought into contact with water, serum, blood and other liquids, if oxygen and carbon dioxide are present, dissociation occurs and a substance is formed, probably a hydrated basic carbonate of nickel, which is slightly soluble, but also forms a very fine precipitate, which renders the liquid turbid. The liquid can only be cleared of the turbidity by passing it through a Berkefeld filter.

6. Water dissolves about 0·0035% of the product of dissociated nickel carbonyl (reckoned as nickel) at 18° C. About the same amount is dissolved in 10% solutions of sodium carbonate.

The solubility of this product of dissociated nickel carbonyl is greater in serum than in water. Serum dissolves about 0·0183% (reckoned as nickel). A solution of the salts of serum, under the same conditions, dissolves 0·005% (reckoned as nickel). Solutions of sodium phosphate of approximately the same strength as the phosphates in serum, dissolve 0·0025% (reckoned as nickel).

7. Nickel hydrate is insoluble in water but is soluble in serum to the extent of 0·0225 gr. nickel per 100 grs. serum, and in sodium phosphate solutions to the extent of 0·012% (reckoned as nickel).

8. Nickel carbonate is soluble in water to the extent of 0·0005% (reckoned as nickel), and in serum to an amount, corresponding to 0·016 gr.% of nickel. Salts of serum dissolve 0·0045 gr. % of nickel carbonate reckoned as nickel. Nickel carbonate is insoluble in sodium phosphate solutions.

9. Dissociation product in aqueous solution is precipitated by boiling, and to a certain extent by the addition of electrolytes. Some of it is thrown out of solution spontaneously on standing for some days, whereas solutions of nickel carbonate are not precipitated by either of the above means.

10. Solutions of dissociation product dialyse more slowly than solutions of similar strength of nickel carbonate, and at a certain stage in the dialysis the greater part of the nickel is adsorbed into the membrane.

11. No constant ratio exists between the nickel and the nitrogen contents of solutions of dissociation product in serum of different strengths.

12. When the proteids of the serum, containing the product of dissociated nickel carbonyl in solution, are precipitated by ammonium sulphate, nine-tenths of the nickel remains in solution. One-tenth is precipitated by adding sufficient ammonium sulphate to produce one-third of saturation, and about an additional one-hundredth by producing one-half of saturation.

13. The product of dissociated nickel carbonyl when dissolved in serum is only incompletely removable from solution by dialysis.

The product of dissociated nickel carbonyl is therefore not identical with nickel carbonate or nickel hydrate, and appears to exist in a condition of colloidal solution.

14. The poisonous properties of nickel carbonyl do not depend, as was at first supposed, on the carbon monoxide of the compound.

15. Nickel carbonyl when mixed with air and inhaled by an animal, cannot be absorbed as such, as it becomes split up into the nickel containing substance (? hydrated basic carbonate of nickel) and carbon monoxide, before or soon after reaching the alveoli of the lungs.

16. The poisonous effects of nickel carbonyl are entirely due to the nickel of the compound. The peculiar toxicity of the compound is due to the fact that being introduced in a gaseous form and that the nickel is deposited as a slightly soluble compound in a very fine state of subdivision over the immense area of the respiratory surface.

Various species of the Order Sarcosporidia, some of which have been studied with great care, infest the muscles of different kinds of birds and mammals. Muscle parasites resembling the one described below do not, however, appear to have been described.

1. Monkeys are susceptible to yaws. The skin eruption in the monkeys I have experimented with (Semnopithecus priamus and Macacus pileatus) is, as a rule, confined to the seat of inoculation but the infection is general, as is proved by the presence of the Spirochaeta pertenuis in the spleen and lymphatic glands.

2. Material obtained from persons suffering from yaws and apparently containing Spirochaeta pertenuis only is infective to monkeys.

3. When the Spirochaeta pertenuis has been removed from this material by filtration, the latter becomes inert.

4. The inoculation of blood from the general circulation and blood taken from the spleen of yaws patients into monkeys may give positive results.

5. The inoculation of the cerebro-spinal fluid of yaws patients gives negative results.

6. Monkeys successfully inoculated with yaws do not become immune for syphilis.

7. Monkeys successfully inoculated with syphilis do not become immune for yaws.

8. By means of the Bordet-Gengou reaction, it is possible to detect specific yaws antibodies and antigen.

9. Yaws antibodies and antigen are entirely different from syphilitic antibodies and antigen.

10. The presence of the Spirochaeta pertenuis in monkeys experimentally inoculated, as well as in yaws patients, is practically constant in the unbroken eruptive lesions; the Spirochaeta is frequently present in the spleen and lymphatic glands.

11. Yaws is generally conveyed by actual contact, but under certain circumstances it may be conveyed by flies and possibly by other insects.

The outbreak of cattle plague in Egypt and the establishment of the Serum Institute at Abbassieh offered a suitable opportunity for investigating some points concerning the nature of the causal agent of the disease, and especially the important question whether the virus is, or is not, capable of traversing a filter candle. On this point the literature is strangely contradictory.

1. The high estimate formed by Böhme of the utility of this test has been fully confirmed.

2. Other substances, such as skatole-carboxylic acid, which tended to confuse the results of the older method, do not give an indole-like reaction with this reagent.

3. The test is more delicate, and the time required to ascertain the presence of indole is shortened.

4. The method lends itself to accurate colorimetric quantitative estimations, and the amount of indole present at various stages during the cultivation of certain organisms has been determined.

In the foregoing paper (p. 512) relationships have been established between loss of weight in test guinea-pigs and dose of diphtheria toxin injected. In connection with the above, the normal increase of weight during the period of observation, the effect of the animal ceasing to eat, or any change of diet, had to be taken into consideration. The following note records observations on normal increase of weight, with constant and variable diet, on healthy animals.

The basis of this paper is the analysis of the bacteriological examination of the fauces of a series of 1019 patients admitted to hospital as cases of scarlatina and a consideration of their subsequent clinical histories. A practice was made of swabbing the fauces of scarlet fever cases without any selection on their admission to hospital. In all cases one bacteriological examination was made and in a number of the cases the examination was repeated.

It has long been well established that during the height of scarlet fever streptococci flourish and become greatly predominant in the pharyngeal and buccal mucus. To quote only a few of the many recorded observations.

1. There were 90 patients at the Eastern Hospital, who, during the years 1897, 1898, 1899, 1904, 1905 and 1906, were injected with horseserum for a second attack or relapse of diphtheria, who had been treated with horse-serum at the first attack.

2. Of these 9 had an “immediate reaction,” 22 an “accelerated reaction,” and 8 had both, 39 in all, or 43·4 per cent.

3. The “immediate reaction” showed itself a few minutes to six hours after the first injection for the second attack.

4. The shortest interval between the first injection at the primary attack and the occurrence of an “immediate reaction” after a reinjection for a second attack was 35 days; the longest was 363 days.

5. The ”accelerated reaction” showed itself 19 hours to five days after the first injection for the second attack.

6. The shortest interval between the first injection at the primary attack and the occurrence of an “accelerated reaction” after a reinjection for a second attack was 25 days; the longest was 1511 days. In each of these cases the reaction occurred the day after the reinjection.

7. In seven of the cases of “immediate reaction” the symptoms were severe. In all save one of those of “accelerated reaction” they were mild or moderate.

8. In 24 of the cases there was a normal serum-reaction after the reinjection for the second attack; but even this occurred in most cases slightly earlier than is usual after injection at a primary attack.

9. So far as the figures go they show:

(a) that when a patient has had a serum-reaction at the primary attack, he is more likely to have an abnormal reaction after injection at the second attack;

(b) that the greater the volume of serum given in the primary attack, the more likely is an abnormal reaction to occur after serum at the second attack;

(c) and that consequently an abnormal reaction is most likely to occur after serum at a second attack when the patient has had a large volume of serum followed by a reaction at the first attack.

There are, however, indications that the volume of serum given at the primary attack may have different effects as regards the determination of an “immediate” or an “accelerated reaction,” according to its amount.