114 
MESSRS. J. N. LANGLEY AND H. M. FLETCHER 
If we take here the mean rate of secretion, we see that there are several divergences 
from Heidenhain’s law ; of these I. and VI, can be explained by taking into considera¬ 
tion the variation in the rate of secretion during the time of collecting I. In I. the 
4 c.c. of the saliva collected is secreted at a rate of ‘9 c.c. a minute, and 1 c.c. at a 
rate of ‘5 c.c. a minute. In V. the percentage of salt is '472 m saliva secreted at a 
rate of '400 c.c. a minute, so that, allowing an increase of ’004 per cent, for each 
'01 c.c. increase in rate, the percentage of salts in saliva secreted at a rate of '5 c.c. a 
minute would be ‘512. Hence, the saliva in I., secreted at a rate of ‘9 c.c. a minute, 
contains '616 per cent, of salts, i.e., a higher percentage of salts than sample VI. 
On the other hand, the difference in the percentage of salts between II. and III. 
cannot be altogether satisfactorily explained in this manner. The saliva III., secreted 
at a slower rate than II., should have a lower percentage of salts; in fact, the per¬ 
centage of salts in it is higher. It is true that a portion of III. may have been 
secreted at a faster rate than any in II,, for in III. 4’8 c.c. were secreted at a mean 
rate of 1'2 c.c., and, as the rate of secretion slackened towards the end of stimulation, 
the rate was, of course, considerably faster than 1'2 c.c. a minute at the beginning of 
the stimulation; but, in view of the slight increase in the percentage of salts which 
occurs as the rate of saliva approaches its maximum (cf. p. 117), this explanation is 
insufficient. It is possible that the increased percentage of salts may have been due 
to the blood-flow through the gland being in this case less than normal [cf. p. 12G). 
In the experiment given above, the stimulation of the chorda tympani was, in most 
cases, stopped as soon as the secretion, beginning fairly rapidly, began to be obviously 
slower. In the following experiment, a variation of this procedure was made by 
stimulating the chorda for a definite short portion of each minute, so that the stimulus 
ceased at about the period of the maximum rate of secretion for the stimulus used. 
With both methods, there is usually a small amount of saliva secreted slowly, after the 
stimulation of the nerve has ceased. 
