^8 - Repoki S.A.A. Advancemkni oI' Science. 



contain Pheoganeine, the two other colouring matters being decom- 

 posed by a slow reaction. 



Some old yellow wines contain a small quantity of Rhodoganeine. 

 If acid is addetl- if they are mixed with soda water, for instance — the 

 colour of Rhodoganeine is intensified, • and becomes apparent. If 

 the acid is saturated by an alkali, the wine resumes its former colour. 



If we consider that " casse " does not modify the percentage of 

 acidity (Lagatu) and if we admit that it is caused by the black matter, 

 we must conclude that the black matter is without action on the 

 Pheoganeine, but acts on the two other colouring matters. 



HOW IS J HE BLACK MATTER FORMED? 



From the above experiments we max consider two kinds of 

 ■" casse." 



(. "Casse ' produced in wines spontaneously, the cause being 

 unknown. We shall call it natural. 



2. " Casse " produced by mixing broken sterilized wine in sound 

 sterilized wine. We shall call it experimental. 



Numerous analyses of wines easily broken show that they are 

 all deficient in tannin and rich in albumenoids. Both these results 

 are explained one by the other, tannin having the property of precipi- 

 tating albumen. " Casse " would, therefore, appear to be due to an 

 •excess of albumenoid matters in the wine, in the first instance. This 

 view seems confirmed by the fact that as " casse " increases the pro 

 portion of black matter increa.ses. while the quantity of albumen 

 ■diminishes. 



We must conclude that the black matter Is formed at the expense 

 ■of the albumen. 



We saw how the black matter was measured. The decrease of 

 albumen was ascertained in the following way : — 



Five litres of affected wine are evaporated in a vacuum. The 

 residuum dissolved in pure hydrochloric acid. The solution reduced 

 to I litre by evaporation. lo grammes are saturated with oxide of 

 silver, filtered and tested with Millon's reagent. The intensity of 

 the colour allows an approximate estimation of the proportion of 

 albumen. 



We tried to produce natural break in a wine artificially albumin- 

 ated. Introducing albumen in the wine was found useless, as it 

 simply acted as a fining. We required the albumen in solution. 

 This was obtained by adding to a must rich in glucose an albumen 

 solution before fermentation. The resulting wine was rich in 

 albumenoid matters, and broke very easily. When broken wine was 

 added to it, it showed, after a few days, the phenomena of " casse " 

 in an exaggerated form, with decrease of albumen and formation of 

 black matter in greater quantity than that introduced with the affected 

 wine. 



After our first experiment we concluded that the discolouration 

 was not attributable to a chemical action alone. To further ascer- 

 tain whether this was the correct view, we repeated our first experi- 



