SUGAR AND ACID IN GRAPES DURING RIPENING. 27 
In Table 6 it will be seen that the water-extract sample was used 
for a partial ash analysis as well as for the organic work. The pur- 
pose in view was to ascertain whether the thorough extraction in 
boiling water to which the fruit was subjected would yield an ash 
which gave alkalinities like or different from the whole fruit when 
burned. Several analyses were made to ascertain the comparative 
content of phosphoric acid in the extract and in the whole fruit. 
The results differ so widely, both for alkalinities and phosphoric acid, 
that they do not support each other; that is, if the organic salts were 
calculated from the alkalinities of the whole fruit ashed, this would 
give results differing so widely as to be impossible of explanation, 
thus raising the question of correct methods of treating the sample. 
From careful checks made of our methods, however, by use of syn- 
thetic solutions the authors are convinced of the reliability of the 
results given by the water-extraction sample. It is possible that 
in ashing the whole fruit it was heated in such manner as to lose a 
portion of the alkali carbonates. The discrepancy is more marked 
as the fruit matures. 
The fact that the attempt to determine chlorin in the water- 
extract sample gave results entirely too high has been noted in 
another section. The results for chlorin in the ashed samples are 
reasonably constant and have an important bearing on the occurrence 
of this element in supposedly sophisticated grape products, which 
will be discussed in a report on the experimental wines. 
The results for potassium, sodium, calcium, and magnesium oxids 
are presented simply as data on the ash composition of pure and 
adulterated grape products, the application of which must await 
further investigation. 
COMPARISON OF RESULTS ON JUICE AND WHOLE-FRUIT SAMPLES. 
The results for the organic elements and the acid salts in the whole 
fruit samples are more readily compared from Table 7. 
The figures given in this table are percentage results for weighed 
samples of whole fruit and percentage results for the 1912 juice 
samples calculated from Table 1. It will be noted that the results 
for sugar as a soluble constituent are noticeably less on the whole 
fruit samples. This is explained by the fact that the seeds, skins, 
and woody fiber of the fruit is a part of the whole-fruit sample, and 
results in lowering the percentage content of sugar because the juice 
extracted by pressure always carries a greater percentage of sugar 
than is found in the residue left as marc or pomace. 
