which has but little influence on conductance. After about an hour regurgitation of pancreatic juice and bile begins as indicated by the tryptic values, also by the color of samples withdrawn. Both acidity and conductance fall sharply, rise due to secondary gastric secretion and again fall due to further regurgitation. The last specimen is very high in trypsin and is nearly pure, pancreatic juice. Here again may be noted the rise of conductance relative to free acidity, so that the final sample

has hardly any free acid, but has a conductance of 165.

Figure 11 shows somewhat similar conditions. Here, however, the tryptic index remains very low and the intense green color which the samples developed on standing showed the regurgitated fluid to be almost entirely bile. The reduction of acidity is more gradual than in the preceding case and the conductance falls off less rapidly. Pepsin, as in the preceding case, remains nearly constant in spite of the dilution of the gastric juice by regurgitation. The reason for this is not entirely clear. Certain results

of this type indicate that the acidity lowering may be due in part to the slightly alkaline duodenal and pyloric secretions which contain pepsin. It has not been shown however that the pepsin contents of these fluids are sufficiently high to explain some of the results obtained. In the absence

of regurgitation the pep

sin rises with the acid, showing a certain parallelism. Regurgitaion brings about a neutralization of acid, while pepsin is not destroyed. Hence the latter will be expected to rise relatively.

The organic acid of the fruit ingested in this case causes titration values for acidity to lie considerably above the true curve for free hydrochloric acid which would more nearly fol

low the curve for conductance. This illustrates the value of conductance determinations as a check on free acid where organic acid is present. The experiment illustrated in figure 12 is unusual in certain respects. Here we have very marked regurgitation of pancreatic juice and extreme changes in acidity with but little influence on conductance. In this case during the first hour and a half the acidity (even a part of that titrating as free hydrochloric acid) is due to organic and "combined" acids which are low in conductance. The conductance is so low in fact that the values obtained are but slightly above those for pure pancreatic juice and bile. Hence admixture with these secretions produces only slight lowering

become markedly reduced, however, the relation of trypsin to conductance becomes inverse because of the lower conductance of pancreatic juice.

In this case no water was taken and the solution entering the stomach was salivary in character and hence relatively high in phosphate. This accounts for the marked difference between free and total acid which we regularly find in such cases. The conductance of the alkali phosphates being low, the conductance curve more nearly follows that for free hydrochloric acid. Several samples of saliva tested by us have shown a conductance of 50-60 × 10-4 at 37°. This is appreciably lower than values given in the literature although filtered saliva may have been employed in earlier determinations.

SUMMARY AND CONCLUSIONS

A retention stomach tube in the form of an electrolytic cell has been devised which makes possible the determination of intragastric conductances at any desired interval of time without disturbance or removal of gastric contents. It may also be used in place of dip electrodes. The tip contains a thermocouple which makes possible intragastric temperature determinations and corrections, and an aspiration tube by means of which samples of gastric contents may, if desired, be collected for analysis.

By means of this apparatus intragastric conductance variations were studied in connection with determinations of total acidity, free hydrochloric acid, pepsin and trypsin. The conductance of gastric juice is mainly due to the free hydrochloric acid which it contains and the same is usually true of the gastric contents.

After the introduction of water or solutions (as sugar solutions) of very low conductance, the curve for conductance very closely follows the curve for free and total acid. This indicates that the equalization of osmotic concentration is brought about primarily by secretion of normal gastric juice.

After the ingestion of food containing protein the conductance curve usually lies below that for free hydrochloric acid as determined by titration because the latter values are high due to gradual dissociation of the protein salt. In the presence of weak organic acid as after fruit ingestion or of phosphate as where much saliva is swallowed, the conductance falls below titration values and is a better measure of free hydrochloric acid.

Aside from the swallowing of saliva, the conductance of which is low, intragastric conductance is, after the first hour or so of digestion, almost always considerably modified by the regurgitation of pancreatic juice or bile or both and possibly to a lesser extent by pyloric and duodenal secretions. The conductance of pancreatic juice and bile being usually very low as compared with that of the gastric contents at maximum acidity, regurgitation tends to markedly lower intragastric conductance as well as acidity. Conductance, however, rises relative to free hydrochloric acid on account of the higher salt content of these regurgitated secretions. After the ingestion of mineral acid, neutralization is brought about in the same manner as during digestion.

In achylias where intragastric digestion is mainly pancreatic in character the conductance was found to parallel the concentration of pancreatic juice as measured by the tryptic index.

Studies of intragastric digestion by this method as well as of the influence of salt solutions in the stomach and upper intestine, are being continued.

The author is indebted to Dr. P. B. Hawk, Dr. C. A. Smith and Mr. R. J. Miller for the privilege of using certain of these cases and data. He desires also to thank Messrs. H. S. Sargent and E. L. Small for assistance in enzyme determinations.

BIBLIOGRAPHY

(1) ENGELMANN: Münch. med. Wochenschr., 1903, li, 41.

(2) FARKAS AND SCIPIADES: Arch. f. d. gesammt. physiol., 1903, xeviii, 551.

(3) BICKEL: Münch. med. Wochenschr., 1904, lii, 642.

(4) MCCLENDON: This Journal, 1915, xxxviii, 180 and 191.

(5) STENGEL and HOPKINS: Amer. Journ. Med. Sci., 1917, cliii, 101.

(6) HAWK: Practical physiological chemistry, Philadelphia, 5th ed.

(7) SPENCER: Journ. Biol. Chem., 1915, xxi, 165.

AN AIR-TIGHT PLEURA CANNULA

S. J. MELTZER

From the Department of Physiology and Pharmacology of The Rockefeller Institute

Received for publication October 1, 1917

In the beginning of the nineties I devised an air-tight pleura cannula. The late Prof. Hugo Kronecker who in 1893 saw the efficient working of the cannula in my private laboratory asked me to give him a brief description of it; this sketch he published later in two journals (1). At the tenth annual meeting of the American Physiological Society (2) I showed the cannula in connection with another demonstration, but published no description of the cannula itself.

About fifteen years ago, I made an essential change in the construction of the cannula. It is the latter construction which we have been using in our laboratory since 1904. The cannula was often mentioned. in papers which emanated from our department, but we never published a description of its construction. In connection with the following paper of Dr. A. L. Meyer, in whose investigation the cannulas played an essential part, I decided to publish the following brief description of our pleura cannula in its present form.

Figure 1 presents the later type of pleura cannula when all its parts are connected; the rubber gasket and two leather washers are here omitted.

Figure 2 shows the four parts composing the pleura cannula.

Part 1. When the larger branches are put together, the entire parts presents a T-shaped hollow cylinder terminating in two flat plates (feet). The vertical cylinder carries on about three-fourths of its length a spiral ridge (thread). The feet which taper toward the beveled end are flat on the lower surface and slightly rounded on the upper side. Both halves of the cylinder hinge at the heel. When they are separated to 90 degrees the entire part assumes again a T-shape in which both feet form one branch that stands perpendicular to the horizontal lines formed by the halves of the cylinder.

Part 2 is a rosette shaped plate with an opening in the middle which is slightly larger than the diameter of the cylinder of part 1. Around