545 



147 cm.'; and in this liquid the endpoint was deternuned. This 

 precaution was omitted, when tiie total volume was between 140 

 and 147 cm.' at the end of the experiment. This measure purposes 

 to prevent that in an inquiry into the titratable calcium in salt 

 solutions of the same molecular concentration, these would have 

 different molecular concentration at the endjioint of the titration, 

 and would no longer be comparable for this reason. 



1/10 N neutral solutions of the sodium salts were made from 

 citric acid and its various substitution pioducts (neutral towards 

 litmus; it was verified that they remained neutral towards litmus 

 on dilution with the same volume of the above gypsum solution). 

 As normal solutions were considered those that contained one gramme- 

 molecule per litre (hence not: One gramme-equivalent in multi-basic 

 salts). The mixture of gypsum and of (perhaps substituted) citrate 

 accordingly contained the various salts in the concentration of 1/20 N. 



The gypsum solution diluted with the same volume of water 

 consumed on an average 45.7 cm'. This corresponds with 12.2 parts 

 of CaO per 100000 parts of water; or with 8.7 parts of Ca per 

 100000 parts of water. In citrates etc. it was derived from a table 

 of Lunge and Buhl ') (calculated from experiments by Faist and 

 Knauss), how much Ca was mo/ found back in the titration, calculated 

 as percentage of the total quantity (8.7). 



In the first column is given the consumed quantity of cm' of 

 titration liquid; in the second column the quantity of calcium that 

 was not found back as percentage of the total quantitj'. 



Thus I found: 



a. Citric acid *) 



b. Te alcohol group made inactive. 

 Acetylcitric acid 



Compared with : 

 Aconitic acid 

 Tricarballylic acid 

 Isoallylene tetra carbonic acid 



c. One carboxyl group made inactive. 

 Symmetrical citric acid monoamlde 



Compared with: 

 Apple acid 

 Tartaric acid 



') 6th edition, Vol. 11, p. 232. 



^) When so few cm* of titration liquid are sutBcienl to reach the limiting value, 

 the limit is much less easy to determine than it is otherwise, and the observations 

 differ much more from each other. 



36 



Proceedings Royal Acad. Amsterdam. Vol. XXVI. 



