48 PHYSIOLOGICAL CHEMISTRY. 



Some authors have gone so far as to recognize several modi- 

 fications of achroodextrin, which are described as a, and y 

 forms, and which differ in optical properties and reducing 

 power. The more recent extended investigations seem to dis- 

 prove this notion, however, and the most that can be safely 

 said is that along with maltose an end product is produced by 

 diastatic conversion of starch which is probably a single sub- 

 stance. What is called erythrodextrin is more likely to be a 

 mixture, possibly, of soluble starch and achroodextrin. Under 

 the name erythrogranulose a similar complex has been de- 

 scribed. In a later chapter on the action of ferments more 

 will be said about the theory of the transformation of starch 

 into these products. 



The true dextrins are not directly fermentable with yeast, 

 but they appear to be aldehyde bodies and as such have 

 reducing power. They react also with phenyl hydrazine and 

 yield osazones, which, however, are not easily purified, because 

 of their solubility. The dextrins have a slightly sweetish taste 

 and all show a specific rotation about [a] n + 196. Be- 

 yond the empirical formula C 6 H 10 O 5 it is not possible to go in 

 describing the constitution of these bodies. 



The natural vegetable gums are often mixtures of several 

 substances, and but few of them have been studied. Gum 

 arabic and gum Senegal are the potassium and calcium salts of 

 arabic acid to which the formula (C 6 H 10 O 5 ) 2 + H 2 O is given. 

 On treatment with weak sulphuric acid both arabinose and 

 galactose appear to be formed. Agar-agar is said to yield lac- 

 tose and then galactose, while cherry gum yields arabinose. 



Cellulose. The cell walls of vegetable substances consist 

 of cellulose mixed always with related compounds of which 

 a body called lignin is the most important. The cellulose 

 resists the action of strong oxidizing or other agents much 

 more perfectly than do the accompanying bodies, and may 

 therefore be freed from them by various treatments. In 

 washed Swedish filter paper we have an illustration of nearly 



