PHYSIOLOGICAL ACTIONS OF DRUGS ON THE SECRETION OF BILE, 139 
biliary flow is rendered somewhat unequal by irregular diaphragmatic com- 
pression of the liver. Moreover, if the respiration be deficient, the secretion 
of bile is always diminished. Some of the slight oscillations observable in the 
charts of the biliary secretion in these experiments are probably owing to varia- 
tions in the respiration ; for in the earlier experiments we were obliged to 
have the respiratory bellows moved by the hand, and this is never so regular 
asamachine. Notwithstanding this, however, the main results of these experi- 
ments are perfectly clear. 
Until it is attempted, one might suppose that this mode of experiment is 
extremely simple, but it is by no means so simple as it appears. It is needful 
to manipulate the abdominal viscera with great care, and to avoid all dragging 
at the bile-duct, otherwise the secretion of bile becomes so irregular that the 
experiment may be useless. The cannula must be very carefully retained in a 
position which will permit of its moving with the diaphragm, but will prevent 
it from twisting the duct, and thus impeding the exit of the bile by forming a 
valve at its orifice. 
Roéurie estimated the velocity of the biliary secretion by counting the 
seconds that elapsed between the fall of the drops from the orifice of the tube. 
A single trial convinced us that this method is extremely laborious, and leads 
to inaccurate results, because it does not permit of continuous observation for 
any length of time. Variations in secretion often occur independently of the 
administration of any substance, and it is impossible to estimate their signifi- 
cance, and make due allowance for them, unless the method of continuous 
collection of the bile be adopted. Moreover, we saw that the degree of 
viscosity of the bile caused a variation in the size of the drops, and, therefore, 
in the intervals between their fall. We therefore abandoned this for the more 
accurate method of allowing the bile to flow into a fine cubic centimetre 
measure, and recording the quantity secreted every quarter of an hour. In 
addition to constant collection of the bile, this method has the great advantage 
of permitting a graphic representation of the results. 
It is evident from the method of experiment that all our observations 
relate exclusively to the effects of substances on the 0zle-secreting mechanism. 
We have made no observations regarding their effects on the bile-expelling 
mechanism. Nor do we intend to prosecute the latter part of the inquiry, for 
the question, what substances stimulate the liver to secrete more bile, is of 
infinitely greater importance. We shall be able to give to it a precise answer, 
and thus for the first time to furnish the physician with definite knowledge for 
his guidance in the treatment of hepatic disorder. 
In several instances we analysed the bile secreted before and after the 
administration of a drug, but although valuable facts were thus ascertained, 
we found that in consequence of the excessive labour of this research it was 
