The acidity of the "purest" juice known at that time was scarcely 0.3 per cent. As a further proof that none of the older observers ever really obtained a secretion from mechanical stimulation pure and simple, we may adduce the fact that none of them made mention of the constant and precise period of five minutes' latency. To overlook this was not possible if a genuine excitation of the glands had been obtained.

Of no less importance is the second condition when we wish to perform the experiment of mechanical stimulation in the correct way. It is very necessary that the gastric glands be not already in activity at the beginning of the experiment, and also that during the experiment no impulse comes into play, which of itself, apart from mechanical excitation, could excite the glands to secretion. Nor have we any proof that observers formerly waited for hours before commencing the experiment and convinced themselves that the gastric glands had ceased working. On the contrary, we have not the slightest evidence to indicate that the authors had attempted to guard against accidental psychical stimulation of the glands - a matter which we have seen is of considerable difficulty. And some dogs are so easily excited in this way that it is almost impossible to bring their glands to rest, or at least it is necessary to wait for hours. The experimenter must strain his whole attention to preserve such an experiment free from objection.

It is only necessary that some food be near the dog, or that the hands of the attendant who has prepared the food should smell of it, or that some other similar circumstance should come into play, and the glass tube, quite undeservedly, will be made answerable for the excitation of the gastric glands. As you have just seen, both of our conditions have been fulfilled on the dog before you, and the result of the experiment stands in irreconcilable contradiction to those of the laboratory and lecture experiment of former times.

The importance of the experiment, which I have already dwelt upon, justifies me in making still further demands upon your attention in order to show you two modifications of it. Nobody has as yet said, with regard to mechanical stimulation, that in order to obtain results the mechanical agency must simultaneously come into contact with numerous points of the inner surface of the stomach. But in order to meet this possible objection I will now show you two new modifications. Again a similar dog is used, that is to say, one on which both gastrot-omy and cesophagotomy have been performed. The stomach has been washed out clean and is at present in a state of complete rest. Into the fistula I bring a thick glass tube containing a number of small openings (2 to 3 mm. diameter) at its rounded end. The other end of the tube is connected with a glass ball containing tolerably coarse sand. Leading into the ball is a second tube, with which an india-rubber pump can be connected and a blast of sand blown through. By rhythmic compression of the india-rubber ball I inject sand with considerable force into the stomach, and this play is kept up for ten to fifteen minutes; nevertheless, we see no trace of gastric juice.

The sand falls out again between the side of the cannula and the glass tube, and it is either dry or scarcely moistened, but in no case is it able to turn blue litmus red. And yet we are here dealing with a strong and widely diffused stimulus. Look for a moment at the performance of the bellows outside the stomach. From every opening of the tube -numbering considerably more than ten - a strong stream of sand is ejected. If you hold your hand against it, you feel quite distinctly that the grains of sand strike with considerable force. And now, when our experiment is ended, we may convince ourselves by sham feeding, in easy and unquestionable fashion, that the innervation of the dog's stomach is perfectly normal. Yet another experiment on a similar dog. Into its empty and resting stomach an india-rubber ball is introduced. This is distended with air by means of a syringe till it is as large as a child's head and maintained in this condition for a time, afterwards being allowed to collapse. The procedure is kept up for ten to fifteen minutes. During this time not a single drop of juice has appeared from the stomach. The surface of the ball taken out of the organ is everywhere alkaline. And here also subsequent sham feeding shows that the dog is in a suitable condition for the experiment.

I must add that in making this observation the dog must not be too hungry, that is to say, must have been fed within ten to twelve hours before, otherwise a psychic excitation of the secretion can readily be induced.

If one dispassionately regards this question, and if any of our methods for the study of gastric secretion are reliable, one must be convinced step by step in the laboratory of the uselessness of mechanical stimulation. In the case of dogs with an ordinary gastric fistula, and failing some special reason, not a drop of gastric juice ever escapes from the stomach other than during the digestive period. How could this be the case if the mechanical stimulus were effective, since the inner rim of the fistula-tube is continuously in contact with the gastric mucous membrane? The same holds good for the dog with resected stomach. During the experiment a glass or india-rubber tube is brought sufficiently far into the cul-de-sac to catch the juice, and yet not a drop flows through the tube, nor does its inner surface ever become acid, so long as true secretory conditions are absent. Moreover, the tube has tolerably often to be taken out and set right.