186 
Philippine Journal of Science 
1920 
was proven not to be the case, as shown by the difference in their 
protein distribution. The results with the 100-mesh sample were 
still low. The method was then slightly modified, using the 
sample employed in the second trial. The mixture of sample 
and solvent was shaken repeatedly in a shaking machine, filtered 
and washed until the washing no longer gave a reaction for 
protein. Twelve hours were found to be sufficient to finish the 
extraction. This trial, although decidedly better than the first 
two, extracted only 47.55 per cent of the total protein. Since 
this laboratory has no facilities for very fine grinding, it was 
decided to discontinue the work for the present, and to report 
the results obtained. 
A comparison of the results of trials 1 and 2 shows that the 
two samples, one composed of the materials passed through the 
40- to 100-mesh sieves, and the other composed of the portion 
passed through the 100-mesh sieve, were not the same in general 
composition. The difference in protein distribution may be due 
to the fact that the sample used in trial 2 was composed of that 
portion of the copra meal which could be pulverized more easily ; 
and it is not surprising to find, as in this case, that its 
composition was different from that of the portion harder to 
pulverize. 
The results of trial 3 show that the modified method used is 
better for copra meal than the one proposed by H. H. Snyder. 
In this trial, the percentage of salt-soluble protein is lower than 
in trial 2. Most of the water-soluble protein is also salt-soluble ; 
and, therefore, an increase in the water-soluble extract naturally 
causes a decrease in the salt-soluble fraction. When the per- 
centages of the water- and the salt-soluble moieties are com- 
bined, the results in trial 3 show an improvement as compared 
with those in trial 2 
Calculated in percentage of protein that went into solution, 
the mixed sample yielded more water- and salt-soluble than did 
the 100-mesh sample. The water- plus the salt-soluble in trial 
3 is lower than in trial 2 by about 5 per cent in the alkali-soluble 
fraction of the former, which might explain the difference. 
There is practically no alcohol-soluble protein. This partly 
agrees with the observation of Osborne (13) that alcohol-soluble 
protein is not found in any seed but cereals. 
All the trials show that the alkali-soluble protein is the most 
abundant in copra meal. 
Although this work is incomplete, it is believed that it may 
serve as a guide in the study of the biochemistry of copra meal. 
