Dec. 15. 1914 Changes in Composition of Ripening Bananas 1 89 



Nitrogen determinations were made by the official methods' by Mr. 

 T. C. Trescot, of the Bureau of Chemistry. 



Ether extract was determined by exhausting with alcohol, and de- 

 termining the ether extract in residue and alcohol-soluble extract. In case 

 of the alcohol-insoluble extract the alcohol was evaporated, the residue 

 transferred to a separatory funnel by use of water and ether, and extracted 

 repeatedly with ether. These extracts were then combined, washed with 

 water, and evaporated in tared flasks. 



Carbohydrates were determined by treating weighed samples with 

 80 per cent alcohol in Soxhlet extractors, using several portions of alcohol 

 so as to avoid giving the larger proportion of the sugars a prolonged heat 

 treatment. The residues were dried and weighed, and the weighed por- 

 tions used in the estimation of starch and pentosans. The extracts were 

 mixed with a little calcium carbonate, evaporated on the steam bath in 

 a current of air, avoiding evaporation to dryness, and then taken up in 

 water, treated with lead acetate, and made up to a known volume. 

 After filtering, the excess of lead was removed by use of dry sodium 

 oxalate. 



Reducing sugar and sucrose were determined in this solution by 

 applying the copper-reduction method of Munson and Walker,^ before 

 and after inversion.^ 



Starch. To determine starch, 3-gram samples of the alcohol-insoluble 

 material were transferred to 300 c. c. flasks, hydrolyzed with hydro- 

 chloric acid as directed in the official method,^ and the dextrose deter- 

 mined by the copper-reduction method of Munson and Walker.'^ 



Pentosans were determined in the alcohol-insoluble residues by the 

 provisional method of the Association of Official Agricultural Chemists. ° 



Although at the time it was supposed that due precautions had been 

 taken to prevent inversion of sucrose, unfortunately it is not improbable 

 that more or less inversion may have occurred, owing to the failure to 

 heat the alcohol extract to boihng. 



When removed, the bananas were of a bright clear yellow, with a little 

 green showing at the tips. They were entirely free from decay. The 

 fruit surfaces were moist and waxy, indicating that the humidity had been 

 high. During ripening, the calorimeter record showed that the tempera- 

 tures had varied from 18.8° to 20.4° C, average 20.1° C. The oxygen 

 content of the air supplied varied from 20.4 to 10.8 per cent by volume, 

 average 15.5 per cent, probably sufficient for normal ripening. 



The bananas weighed 137.71 kg. when placed in the calorimeter and 

 132.37 kg. when withdrawn, a loss in weight of 3.88 per cent. The 



' Op. cit., p. 7. •* Op. cil., p. 41. 5 Op. cit., p. 241. 



2 Op. cit., p. 241. * Op. cit., p. 53. "Op. cit.. p. 54. 



