512 The Philippine Journal of Science 1928 
PROCEDURE 
The solvents used in this investigation to extract coconut oil 
from copra cake were ether, carbon tetrachloride, benzene, pe- 
troleum ether, chloroform, acetone, absolute alcohol, alcohol 
(95 per cent), and methyl alcohol. Preliminary experiments 
showed that coconut oil dissolves readily in these various sub- 
stances. The solvents were purified by special means when nec- 
essary and distilled several times until they showed a constant 
boiling point. In working with these solvents we employed the 
usual Soxhlet extractors. Two electric heaters containing six 
extractors each were used. Since the boiling point of the sol- 
vents varied from 34 ° to 80 ° the electric heater employed for the 
higher-boiling solvents was arranged to give a somewhat higher 
temperature than the other heater used for the lower-boiling 
solvents. The extraction temperature was regulated so that 
the rate of .extraction was approximately the same for each 
solvent and for each interval of time. Although the various 
solvents ‘boil at different temperatures, by placing pieces of 
asbestos under the extraction flasks the heat can be regulated 
so that each solvent will require approximately the same length 
of time (about twenty-four minutes) to syphon. When the 
extraction was completed for a definite interval of time the 
heating was continued until most of the solvent had distilled into 
the upper part of the apparatus and was just about ready to 
syphon. The heating was then discontinued. The extraction 
fiask which still contained a small amount of solvent was dis- 
connected and placed in an electric oven heated to a temperature 
of about 75°. The flask containing the extracted residue was 
allowed to remain in the oven until all the solvent had ap- 
parently evaporated, after which it was weighed. The flask 
was again heated and weighed several times until a constant 
weight was obtained. Since ether is the solvent that is com- 
monly employed to extract oil from copra cake, we used the 
results obtained with ether as a standard for comparison. 
In all our extraction experiments we used a 2-gram sample 
of material, and the experiments were always performed in 
duplicate. The duplicate extractions for each interval were 
naturally carried out with fresh samples of copra cake. In 
some cases, when the results of our duplicates did not agree 
very closely, we performed several series of duplicate experi- 
ments until we obtained fairly constant results. 
In many series of experiments we varied the intervals of 
time for a particular solvent from three to thirty-six hours. 
