258 AMYLOLYTIC POWER AND CHEMICAL 



that " the alkalinity of saliva depends upon the presence of 

 sodium carbonate. In man and in the dog the percentage of 

 this salt varies from 0.08 to 0.19 per cent." So far as we are 

 aware, however, there is no justification for this statement. 

 In the earlier work from this laboratory * it was stated that 

 the average alkalinity for fifty-one samples of human mixed 

 saliva was 0.08 per cent, " expressed in the form of sodium 

 carbonate." Further, in all the tabulated results contained 

 in that paper, the alkalinity, as measured by titration with 

 standard acid in the presence of cochineal as an indicator, 

 was carefully expressed as " equivalent in NaaCOg," this being 

 done to avoid any positive statement as to the exact cause of 

 the alkalinity. Further, in the oft-quoted work of Werther f 

 the alkalinity of the saliva of the dog was determined by 

 titration with decinormal sulphuric acid with litmus as an 

 indicator: a method which obviously would throw no light 

 upon the cause of the alkalinity. Moreover, in at least some 

 of the tables containing his results the percentage of alka- 

 linity is expressed as "alkalinity calculated as NajjCOg." 



Examination of a large number of samples of human mixed 

 saliva obtained from different individuals at different periods 

 of the day convinces us that, under normal conditions at least, 

 human saliva never contains the least trace of sodium carbon- 

 ate. Toward litmus, lacmoid, etc., human saliva constantly 

 reacts alkaline, but with phenolphthaleln it invariably shows 

 an acid reaction, and a certain amount of a decinormal alkali 

 solution is required to bring out an alkaline reaction with this 

 indicator. Further, phenolphthaleln is an extremely sensitive 

 reagent for sodium carbonate ; a solution containing 0.001 per 

 cent of sodium carbonate will give a pink color when brought 

 in contact with a solution of phenolphthaleln. With human 

 saliva, however, we have never obtained any color reaction 

 with phenolphthaleln whatever; the solution invariably re- 

 mains colorless, thus proving that the alkalinity indicated by 

 litmus must be due to some acid salt or salts, like the hydro- 



* Chittenden and Ely, Amer. Chem. Journ., 1883, ir, p. 333. 

 t Werther, Archiv f. d. ges. Physiol., 1886, xxxriii, p. 293. 



