(1) The graphic method is more easy of application and requires less mathematical knowledge than the extended analytical method.

(2) In our experience it produces smooth curves of px and Ex.

(3) The analytical in the selection of the best process of interpolation appeals to the graphic method.

(4) While either the analytical or graphic process may in unskilful or careless hands give erroneous results, owing to errors in working out, the analytical process, unlike the graphic, presents a wide choice of methods, which although accurately worked out, give incomparable results.

(5) The facts given in the table on page 321 and in Fig. 1 A and B and table on page 317 show that almost identical results are obtained by the graphic method and by the improved extended analytical method. We regard this as important testimony to the accuracy of each of these methods, and at the same time as indicating the inaccuracy of the analytical methods giving different results.

(6) As the results obtained by the modified short analytical method and by the combined analytical and graphic method approximate so closely to those obtained by the detailed graphic method, either of them may safely be adopted by those who cannot spare time for working out a detailed Life Table by the graphic method.

(7) Every published Life Table should give exact details of the method of its construction, in order that the comparability of data and results may be tested. Much confusion has arisen, and important errors in comparison between different Life Tables have been produced, by the use of different methods of construction and the non-publication of details of methods.

The various problems of infantile mortality are not only supremely interesting to the medical officer of health, but, with the continued decline in the birth rate, are calculated to assume even greater practical importance in the birth rate, are calculated to assume even greater practical importace in the near future. Of these problems none is more in need of solution than the exact etiology of infantile diarrhoea.

When the publication of the series of Reports for each separate County, dealing with the facts ascertained at the Census of 1901, was completed last March, it was possible for anyone, who would take the trouble, to arrive at the facts relating to the whole of England and Wales, that is, to ascertain not only the finally corrected total population number, but the numbers classified into age and sex groups.

1. Powerful anti-sera may be produced by the intravenous injection of smaller quantities of serum than have hitherto generally been used (p. 260).

2. Nuttall's quantitative method affords a simple and fairly accurate means of determining the quantity of precipitum formed. By its means quantitative differences can be appreciated which are scarcely, or not at all, apparent in the tubes on inspection (p. 263).

3. Normal saline solution is the best diluent for normal sera, and 1—21 has been found to be a convenient dilution. Increase of salt has very little effect on the production of precipitum (p. 266).

4. The quantity of precipitum formed is not influenced by the temperature at which the experiment is conducted (i.e. between the temperature of the ice-chest and 37° C.) (p. 268).

5. In the case of dried bloods time per se does not destroy their capacity for reacting with their homologous anti-sera. Fluid sera appear to deteriorate slightly by keeping (pp. 269—274).

6. Putrefaction of the serum, or anti-serum, does not affect the production of a specific precipitum (p. 274).

7. Although the intimate mixture of lime and blood completely destroys the latter, the former is not present in sufficient quantity in ordinary earths to affect blood mixed with earth. The presence of small quantities of lime, however, gives rise to a clouding in solution, which can be got rid of by the passage of CO2, and subsequent filtration (pp. 276—281).

8. The presence of even small quantities of acids, or alkalis, rapidly reduces the quantity of precipitum formed (pp. 281—284).

9. In diseased conditions a marked alteration may occur in the quantity of precipitum (pp. 265 and 285).

10. The volatile antiseptics produce little effect on sera, even after long contact, but formalin, corrosive sublimate, lysol, lysoform, the sulphates of copper and iron, and nitrate of silver, especially in strong solutions, exert a very deleterious action (pp. 287 and 358).

11. Blood dried on fabrics, and materials in common use (with the exception of certain leathers) may with adequate precautions be readily diagnosed (p. 290).

12. After an undiluted anti-serum has been raised to a temperature beyond 60° C. the capacity for producing precipitum is diminished, and it is destroyed completely after exposure to 68° C. These effects seem to be produced at lower temperatures in normal undiluted sera (pp. 354–55).

13. The precipitum-producing power of normal sera is reduced by filtration through a Chamberland filter, but not by passage through a Berkefeld filter (as far as the experiment was conducted) (p. 357).

In the first place cultures of the nitroso-bacterium were developed in inorganic solutions. These carried to a 5th dilution exhibited practically a pure culture of this species.

Secondly, cultures of the nitroso-bacterium were inoculated into solutions containing small quantities of organic matter. In these they were able to oxidise the ammonia present. It was found that a culture developing in the presence of small quantities of organic matter was better able to oxidise the ammonia in higher percentages of organic matter than a culture taken directly from an inorganic solution.

In growing this species on plates silica jelly has the disadvantage of being difficult to prepare. It was 18 months before I obtained a satisfactory plate. This jelly grows the nitroso-bacterium well, but other species can also develop colonies on it. Ammonia agar also grows the species, as shown by the oxidation of the ammonia, and the colonies assume a characteristic form.

As to beef broth agar and gelatine, these media grew colonies of a micro-organism similiar to the nitroso-bacterium from oxidised ammonia cultures. Pieces of these plates containing such colonies in pure culture frequently oxidised ammonia in solutions. On the other hand pieces of such plates showing no colonies never produced any oxidations of the ammonia. Furthermore the single colonies from silica and ammonia agar plates, which oxidised the ammonia in the media into which they were sub-cultured, grew well on beef broth agar and gelatine.

From the above results I have come to the following conclusions:

(1) That the nitroso-bacterium grows well on any ordinary medium.

(2) That the supposed parallel organism is no other than the nitroso-bacterium itself.

(3) That in the presence of large percentages of organic matter the nitroso-bacterium, although growing very profusely, loses for a time the power of converting ammonia into a nitrite.

An instrument which I had made several years ago presents some advantages and may therefore be of use to others who have to deal with serum-diagnosis, etc. The chief advantage of the instrument is that after each dilution has been made it is ready for making further dilutions without cleansing it. The instrument is moreover easy to manipulate. It may be described as essentially a “reversed Thoma-Zeiss pipette”, the graduations being marked upon the upper section of the tube, instead of on the lower section. The accompanying figure makes the construction of the instrument clear.

It is matter of common observation that air which is much vitiated by combustion of lighting-gas is distinctly oppressive, apart altogether from the rise of temperature which is always associated with the vitiation. This effect is always very evident if the proportion of CO2 in the air of a room has been raised to 30 or 40 volumes per 10,000 by combustion of gas. Air to which pure carbonic acid has been added in the same proportion has no such effect, however, and is practically indistinguishable from pure air. Deficiency of oxygen to such an extent as occurs in the air of a room is also without sensible effect. In coal-mines, where the air is commonly vitiated to a considerable extent by slow oxidation in the strata, it is, for instance, quite impossible to distinguish by the senses pure air from air containing an excess of 50 or even 100 volumes per 10,000 of CO2, with a correspondingly large deficiency of oxygen. The unpleasantness of air vitiated by combustion of lighting-gas is therefore not due to excess of carbonic or deficiency of oxygen.

In this paper the term B. coli is used in its most restricted sense, and as implying organisms having all the characters of the typical B. coli communis. It is a well-ascertained fact that B. coli, even when the term is so restricted, exhibit very varying virulence.