360 Mary Davies Swartz, 



organism. While we have, in the case of herbivora, some convincing 

 evidence that the pentosans are a true source of energy/ we have as 

 yet no real grounds for this assumption in the case of man. 



The insoluble pentosans of the Hawaiian algae are manifestly less 

 digestible than the soluble forms. The coefficient of digestibility is 

 approximately the same as Slowtzoff's (154) average for pure xylan 

 in rabbits, 55 per cent. While it would be perhaps desirable to de- 

 termine the pentosans directly by the furfurol-phloroglucin method, 

 rather than by estimation of sugar after acid hydrolysis, a trial with 

 dulse by both methods gave practically identical results: hence, con- 

 sidering that the hemicelluloses of these algae are chiefly pentosans, 

 it seems safe to assume that the results reported represent the amount 

 of pentosan excreted, within the limits of error for all of the feeding 

 experiments. 



The galactans were all soluble, and were ingested in quantities not 

 exceeding 15 grams per day, yet the coefiicient of digestibility is lower 

 than for any other hemicellulose group (26 per cent) . The resistance 

 of Irish moss is particularly striking, but is not surprising in view of 

 its utter indifference to attacks of digestive enzymes or bacteria. Its 

 influence on the character of the faeces was not so marked as that of 

 Limu Huna, owing probably to a greater tendency to liquefy at body 

 temperature. The latter would seem to be a very effective agent in 

 constipation; a comparison of its efficiency with that of agar-agar 

 would be extremely interesting. Saiki (205) found the coefficient of 

 digestibility for agar (average of two experiments) 17 per cent. 



In view of the negative results of digestions in vitro and of trials 

 with bacteria, we can scarcely be surprised at the results of these met- 

 abolism experiments, especially as we recall that Lohrisch (57) found 

 that his "soluble agar," already partially hydrolyzed, was only diges- 

 tible to 50 per cent (average). 



The mannans stand in striking contrast to the galactans. In the 

 present studies, 99 per cent of the salep administered has been uti- 

 lized, a result in accordance with Kano and lishima's (255) coefficient 

 of digestibility for the Japanese mannan, Konjaku, 82 per cent. 

 Pure mannan fed to a dog, was excreted the succeeding day, seem- 

 ingly unaltered, since it formed a semi-transparent gelatinous mass in 

 the faeces, from which, later, a rich yield of mannose-hydrazone was 

 obtained. The very different result with salep powder, of which 85 

 per cent was digested by dogs, may perhaps be accounted for by the 



3 Cf. Kellner, p. 274. 



