DIGESTION IN THE MOUTH 749 



that the force with which the secretion is turned out into the ducts 

 of the gland is greater than that represented by the blood pressure. 

 The blood pressure in the capillaries must be considerably lower even 

 with full vascular dilatation than the pressure in the carotid artery. If 

 two manometers be connected, one with the carotid artery and the other 

 with the duct, it will be found that on stimulating the chorda tympani 

 nerve the secretion will be excited, and the mercury will be driven 

 up by the pressure of the secretion in the corresponding manometer 

 until it attains a height which may be double that of the mercury 

 in the manometer connected with the carotid artery, and therefore 

 must be still greater than the pressure in the capillaries of the 

 gland. This experiment, which is easy to repeat, showed the 

 impossibility of the act of secretion being in any way determined by 

 a process of nitration. We have now further evidence that work 

 is done in the production of the salivary secretion, evidence which 

 was not available when Ludwig first carried out the experiment 

 just described. When *a fluid containing salts in solution is filtered 

 through a porous membrane the filtrate has the same content in 

 salts as the original fluid. We can effect a separation of dissolved 

 salts from a fluid by filtration under pressure through some membrane 

 which is impermeable to the salts, e.g. a membrane of copper 

 ferrocyanide a so-called semipermeable membrane. Under these 

 circumstances a very large pressure is necessary in order to cause 

 the filtration of any fluid at all, a pressure which is equal to the 

 osmotic pressure exerted by the substances in solution. Thus if 

 we were filtering a 1 per cent, solution of NaCl through a semi- 

 permeable membrane, we should have to exert a pressure of about 

 seven atmospheres in order to obtain a filtrate free from sodium 

 chloride. To obtain a filtrate containing half the amount of sodium 

 chloride, if such were possible, would therefore need a pressure of 

 about three and a half atmospheres. On comparing the osmotic 

 pressures of saliva and blood respectively and for this purpose we 

 can employ the depression of freezing-point as our index we find 

 that the molecular concentration of saliva, and therefore its osmotic 

 pressure, is always very much less than that of the blood plasma, 

 and may vary between half and three-quarters of the latter. Sup- 

 posing the membrane separating the lumen of the duct from the 

 blood-vessels could be regarded as endowed with the properties of 

 a semipermeable membrane, we should need, in order to effect the 

 separation, a pressure ten to twenty times as large as the arterial 

 blood pressure. We must conclude then that work, both osmotic 

 and mechanical, is performed in the separation of the fluid from 

 the blood and its transference in the form of saliva to the duct. A 

 very simple experiment will suffice to show that this work must be 



