Our knowledge of diarrhoea, its causes and varieties, is far from exact, but a classification is useful, even though it may not carry us far towards a diagnosis in individual cases.

(1) Diarrhoea may arise from the ingestion of irritating, indigestible or decomposing food. The actual cause of the diarrhoea in such cases will be considered later.

(2) It may result from defective gastric digestion. In most examples of gastric disorder there is constipation, and particularly is this the case in hyperacid states. But diarrhoea is sometimes noted as occurring when the secretion of hydrochloric acid is diminished. This diarrhoea of gastric origin is probably due to the entrance into the bowel of material that has not been sufficiently prepared in the stomach, and it is not improbable that this insufficient preparation consists partly in insufficient disinfection of the food. At any rate an important preparatory and protective function of the stomach has failed.

(3) Theoretically, diarrhoea can result from defects in quantity or quality of bile, pancreatic secretion, and succus entericus. The first two of these secretions are considered elsewhere. Succus entericus probably plays a very important part in intestinal digestion, but little is known of the results that follow its failure. Besides its action on cane sugar and maltose, it seems to reinforce all the ferments of pancreatic secretion. Its alkalinity may also be of use, and a decrease in its secretion may thus have ill effects by allowing increased acidity in the small intestine.

(4) Inflammation of the bowel, catarrhal or croupous, and ulceration of any origin are important causes. Generally speaking the diarrhoea is more violent in disease of the colon than in disease of the small intestine.

(5) The effect of an abnormal growth of bacteria in the bowel, with alteration in numbers, virulence or species (as in typhoid fever, cholera and some forms of colitis) and the effect of toxaemia (bacterial as in septicaemia or toxic as in uraemia) may be allowed.

(6) Finally, we may feel sure that in some cases diarrhoea depends on nothing more serious than a perversion of peristalsis, either of central nervous origin, or as a result of an abnormal sensibility of the intrinsic nervous mechanism in the intestinal wall.

If dietetic treatment is to be accurate, some attempt must be made to utilize these theoretical considerations. But there are many difficulties. These causes often interact, they may be conjoined in various groupings, and one may lead to another, so that they defy analysis.'

No doubt diarrhoea can result from chemical or mechanical irritants such as mercury, decomposing meat or a meal of unripe apples. But it is exceedingly likely that even such simple causes produce their effect not directly but through bacterial agency, and that the bacteria normally present in the bowel undergo rapid development upon the introduction of such material. At any rate I think it must be allowed that whenever there is actual inflammation of the intestinal wall the immediate cause is the abnormal growth of bacteria and the products of their activity. It appears from Strasburger's observations that in all forms of diarrhoea the output of bacteria in the stools is greatly increased. The weight (dried) may reach 14 grammes or more per diem, as compared with an average of 8 grammes in health. Consequently the dietetic treatment of diarrhoea is largely bound up with the question of the possibility of influencing bacterial growth in the intestine by alteration and manipulation of food.

Roughly speaking there are two classes of bacteria in the bowel, each of which occupies its own territory. In the small intestine the bacteria are mostly such as produce organic acids (lactic, acetic, and succinic) by their action on carbo-hydrates. The lower two-thirds of the colon hold the bacillus coli communis, and other protein-decomposing organisms. It seems to be established that both classes inhabit the caecum and first part of the colon : at any rate bacterial fermentation of carbo-hydrates has been demonstrated in this intermediate area. Only under pathological conditions do the protein-decomposing bacteria of the colon appear or become obviously active above the ileo-ccecal valve.

As regards the colon it must be remembered that here digestion proper is at an end. I do not think that the character of the food can have any material influence on bacteria of the colon, save indirectly by the choice of a diet leaving little residue and containing no hard or irritating particles.

But as regards the small intestine it is probable that diet has some influence on bacterial growth, though it does not follow that we can use this influence satisfactorily in practice. In consequence of the steady production of acids by bacterial action on carbo-hydrates, the contents of the small intestine remain acid, notwithstanding the constant neutralizing effect of the alkaline succus entericus. It seems probable that this acidity has its use in restraining the growth of putrefactive protein-decomposing organisms, which, normally inhabiting the colon, might otherwise overrun and flourish in the small intestine. It is clear that under abnormal conditions such an invasion by putrefactive bacteria does actually occur. It is probable that by an increase of the carbo-hydrate element in the diet we can increase the degree of this protective acidity in the small intestine. In fact, the stools which are normally alkaline may become faintly acid on a free carbo-hydrate diet. There are cases in which a milk-carbo-hydrate diet based on these considerations should be used. Theory and practice, however, are not always in agreement. And it must be realized that beyond a certain point the increased acidity, with the inevitable development of gas, may itself act as an irritant and lead to increased peristalsis and frequent stools. This is the source of the gas in most cases, when intestinal flatulence is the chief complaint.

In fact, this state of excessive fermentation and acidity in the small intestine can be recognized with some certainty. Diarrhoea, colic and flatulence are its chief features. The term " acid jejunal diarrhoea " has been applied to it by Nothnagel.

It is probable also that a similar excess of acidity of a more complex type may arise through deficiency of the alkaline succus entericus, which is a specialized secretion and is likely to fall off in amount under inflammatory conditions. Schmidt and Stras-burger recognize an "intestinal fermentation dyspepsia" as having such an origin. The symptoms of this condition are obscure pain in the abdomen, fullness and hyperaesthesia of the abdomen, a feeling of malaise, loss of appetite, furred tongue, and frequent loose stools, the faeces being frothy, yellow and acid, but containing no mucus.

It will be seen, therefore, that there are cases of diarrhoea where the fermentation in the small intestine becomes undesirable. Rapid improvement will sometimes occur when milk and carbo-hydrates are dropped and protein-feeding is begun.

It must not be thought, however, that the influence of diet on the intestinal flora turns only on the rise and fall of the activity of the acid-producing bacteria in the small intestine. The conditions are probably far more complex, though it is impossible to trace their effects with accuracy. It is probable that other factors exist, such as the consistence of the food, the amount of fluid taken, the nature and number of the bacteria taken with the food, the value of the upper digestive secretions, and the motor power of stomach and intestine. An example showing the complexity of the problem is given by the undoubted fact that the stools in jaundice are often loose and offensive, and suggest that an abnormal bacterial decomposition of protein is taking place. It is certain that this condition is not due, as was thought at one time, to the loss of some antiseptic property contained in the bile. It is probably due to the fact that the normal pancreatic digestion of protein is in some way interfered with by the excess of fat which in the absence of bile escapes absorption. This interference is possibly of the nature of a mechanical envelopment of protein-particles by fat. At any rate in jaundice with loose stools fat should be discarded from the diet.

I think that in most cases of diarrhoea (except perhaps those of purely nervous origin) dietetic treatment involves (1) a recognition of the cause, (2) a choice between a milk-carbo-hydrate and a protein diet. This choice must be made on the lines indicated above. The decomposition of protein is best met by a milk diet or a milk-carbo-hydrate diet in which milk is the chief article. On the other hand, irritative and inflammatory conditions of the small intestine require a reduction of carbo-hydrates and their replacement by protein and fat. But it must be confessed that in many cases the guides are uncertain and the best treatment must come from experience. Particularly is this the case in the various ulcerative conditions of the bowel.

A pure milk diet is sometimes of use, but there is no reason to believe that it has any influence upon intestinal bacteria, except in so far as the bacterial fermentation of the contained lactose may tend to maintain the acidity of the small intestine. I think its merit lies in the small residue which enters the colon, and it is in diseases of the colon that it finds its chief use. Koumiss and kephir may be similarly used.

Finally, in all forms of diarrhoea it is well to avoid all articles containing cellulose, vegetables, fruit, coarse bread-stuffs and grains, sugar, and meat-extracts.

Though we may draw up a list of the conditions under which diarrhoea occurs, we have no working classification. Certain forms of diarrhoea, however, stand out as having some constant characters.