COTTONSEED MEALS. 21 



In samples Nos. 1630 and 1788 not only do the crude fiber figures 

 sh<>\v a slight tendency to be above the average, but this is also true 

 of the pentosan figures. At the same time the protein and fat figures 

 for these samples, without being markedly low, show a tendency to 

 be below normal. In No. 1788 an excess of cotton fiber was found 

 microscopically. It would, therefore, appear that these two samples 

 contained a somewhat larger amount than normal of hulls or cot- 

 ton fiber. It is only fair to add that a third sample of the same goods 

 as samples 'Nos. 1630 and 1788 had a good average composition. 

 Sample No. 1745 was discussed in the preceding paragraph. 



The figures for pentosans are suspiciously large in only three cases, 

 which have already been discussed. 



Most of the samples examined show the presence of a very small 

 amount of starch. It is probable, however, that no starch was 

 really present, the small amount reported being due to inherent errors 

 in the method of analysis. 



For the sake of uniformity of expression the non-redycing sugars 

 are reported as sucrose. They are probably present, however, almost 

 entirely in the form of raffinose, as shown by Scheibler and Mittel- 

 meiers, Berthelot, 6 Withers and Fraps, c and others. None varies 

 so much from the general average as to appear suspicious, except in 

 the case of sample No. 1745, which has already been discussed. 

 Most of the samples contained no reducing sugars, while a few con- 

 tained a very small amount. 



As a whole, the figures for crude fat are below the various averages 

 shown by the compiled data. It would appear from this that the fat 

 is now more completely removed from the seeds than in the past. 

 Only two fats are so low as to appear suspicious namely, Nos. 1788 

 and 1745. A discussion of both of these samples has already been 

 given. 



Ber. d. chem. Ges., 1889, 22: 3118. 



&0ompt. rend., 1886, 103: 533. 



c North Carolina Agr. Exper. Stat., Bui. No. 179. 



