III. EXPERIMENTAL PART. 



Introduction. 



The foregoing review has emphasized the Kmits of our knowledge, 

 both in regard to the chemical composition of marine algae, and their 

 fate in the alimentary tract of men and animals, as determined by 

 actual measurement of intake and output, and as explained by the ac- 

 tion of bacteria and enzymes in vitro. Ten species of marine algae 

 have, therefore, been made the basis of the present investigations. 

 Eight of them were Hawaiian Limu, obtained, as already stated, 

 through the kindness of Miss Minnie Reed, Science teacher in the Ka- 

 mehameha Boys' School, Honolulu. They were dried in the sun, with 

 the salt water adhering to them, before shipping to America. The 

 other two (dulse and Irish moss) were easily obtained in our Eastern 

 markets. 



That the carbohydrates of algae are chiefly hemicelluloses, is indi- 

 cated by the analyses which have already been made; that in many 

 species, these are to a great extent water-soluble, is also well known. 

 In as much as such soluble forms are thus particularly well adapted 

 for nutrition investigation on account of their freedom from all in- 

 crusting substances, which end to interfere with digestion, the present 

 studies have been confined as far as possible to them. Since it was 

 desirable to study the different groups of hemicelluloses, and man- 

 nans and levulans were not found in the seaweeds in sufficient quanti- 

 ties for metabolism experiments, these were obtained from other 

 sources; a mannan from salep, and a levulan (sinistrin) from squills 

 {Scilla mariiima). 



Other investigators in this laboratory are working on a dextran 

 which would naturally be included here, namely lichenin from Cetra- 

 ria islandica; consequently no experimental studies on this group of 

 hemicelluloses have been made. In considering any classifications 

 of these materials, it must be borne in mind that most of these carbo- 

 hydrates are more or less complex in nature, and can be grouped only 

 with reference to what appears to be the chief constituent in any 

 given case. The following list comprises al the species examined, 

 arranged upon this plan: 



306 



