472 
Philippine Journal of Science 
1919 
organisms, thus affording no data concerning enzymes of mold 
origin. As for the enzymes naturally occurring in coconut 
meat, it seems evident from Table III that neither the soluble 
nor the insoluble enzymes played a large part in the deteriora- 
tion of the oil. A slight effect of the fat-soluble, but not of 
the fat-insoluble, enzymes may be traced, especially in the 
column showing the increase in acidity. Unfortunately the 
enzyme-containing sample must be compared in each case with 
one of higher initial acidity due to sterilization. In two cases 
(6 Od and 6 01) this higher initial acidity seems to have had 
a greater effect than the enzymes (3 Od and 3 01). On the 
whole after two years the sterilized oils were little, if any, 
better than those not sterilized. 
Microorganisms. — It has been shown by Walker 20 that, al- 
though moldy copra is responsible for most of the rancidity in 
commercial coconut oils, and molds will damage a turbid oil, 
yet filtered oils are practically unattacked by microorganisms. 
Any effect of microorganisms shown in this investigation would 
be found in the unsterilized oils and would therefore be included 
in the small enzyme effect previously discussed. But since it 
is unlikely that any mold or bacterial development took place in 
the filtered oils, we are limited to the apparent effect of the 
insoluble enzymes, which was zero. It seems improbable that 
any of the results were affected by microorganisms. 
Moisture. — The samples of dry oil sealed in glass tubes (1 Sd, 
4 Sd, 1 SI, and 4 SI) showed an acidity increase of 0.1 to 0.3 
per cent. Theoretically this hydrolysis required about 2 to 6 
milligrams of water per 30-gram sample. Such small amounts 
of water may have been present in the oil when sealed, or they 
may have been formed by decomposition of substances in the 
oil other than glycerides. When a large excess of water was 
added the acidity increase was 0.4 to 0.7 per cent, and a deodori- 
zation was observed, especially in the samples exposed to light. 
In the open samples moisture conditions did not differ much, 
for all were exposed to normal atmospheric moisture and, on 
the other hand, all added water evaporated. It must have re- 
mained longer in the samples in the dark and assisted in the 
oxidation, as is evident from the iodine numbers. 
Nonfatty material . — Examining Table III for the effect of 
nonfatty material I fail to find any “profound influence.” 
There appears a very slight effect in 3 01B and 6 01B, and 6 
50 Philip. Journ. Sci. 1 (1906) 134 and § A 3 (1908) 127. 
