482 Mr. W. C. M. Lewis : Experimental Investigation 



effects began to be observable, so that no further readings 

 could be taken. These peculiarities are shown in fig. 5. 



Fig-. 5. 



50 



40 



80 



2C 





















f 



ll 

 !■ 



















X 







1 

 i 









V 



e— — * 























:c 









0-35 



0-05 0-1 0-15 0-2 0*25 0*3 



Grams of Saponin in 100 c.c. of solution. 



It seems probable that the gelatinizing effects which have 

 been observed at the air-surface are also effective at the oil- 

 interface. Experiments were carried out to test whether the 

 substance was soluble in the oil. It was found to be 

 insoluble. 



Of the substances enumerated in Table III., it will be seen 

 that caffeine shows the most marked effects in proportion to 

 its concentration. It was therefore selected as a suitable 

 substance wherewith to carry out adsorption determinations. . 



Determination of — for Lajfeine. 



The material w r as obtained from Kahlbaum. Its molecular 

 weight was determined by lowering of the freezing-point of 

 water, and was found to be in agreement with the formula 

 (weight 194). The substance employed is the monohydrate 

 and crystallizes from water in long needles. That caffeine is 

 a good example of a non-electrolyte follows from the value 

 for its dissociation constant determined by "Wood *, namely, 

 4 x 10 -14 , which is of the same order as the constant for water. 



The sample of caffeine was first tested, as in previous cases, 

 for any solubility in the oil. The same drop-number was 

 obtained with the oil filtered from caffeine as for the oil fresh 

 from the stock. Caffeine is therefore insoluble in hydro- 

 carbon oil. 



* Wood, Journ. Chem. Soc. vol. lxxxix. p. 1841 (1906). 



