1907.] Cyanogenesis in Plants. 321 



Further, the tabulated results given above show that an increase in the 

 lsevo-rotatory power of a solution of phaseolunatin, partially hydrolysed by 

 the natural enzyme, is produced by the addition of ammonia. This can only 

 mean that a highly dextro-rotatory substance initially produced by the action 

 of the enzyme is transformed by the ammonia into one of lower dextro- 

 rotatory power. 



It will be seen that, taking the following data into account, 



Specific rotation of phaseolunatin — 27°'4 



„ „ a-dextrose +105° 



/3-dextrose +22° 



„ „ the equilibrium mixture of a 4- /3-dextrose +52 a 5 



these results can only be explained on the assumption that the sugar initially 

 produced by the decomposition of phaseolunatin is a-dextrose, and that the 

 glucoside itself is the a-dextrose ether of acetonecyanohydrin. Phaseolunatin 

 appears to be the only naturally occurring a-glucoside so far known, all the 

 other glucosides so far examined yielding the /3-forms of the sugars on 

 complete hydrolysis by enzymes. 



Nature of the Glucosidolytic Enzymes present in Phaseolus lunatus. 



Since a-dextrose is produced from phaseolunatin by the enzyme preparation 

 obtained as already described from the beans of Phaseolus lunatus, it must 

 be assumed that this preparation contains an a-enzyme and, since the 

 glucoside is also decomposed by the maltase of yeast, it would appear that this 

 a-enzyme may be identical with yeast maltase, or is at any rate of the same 

 type. 



The only other possible view is that the a-enzyme of Phaseolus lunatus may 

 be identical with the invertase of yeast, but Fischer has shown that the 

 latter does not decompose a-methyl glucoside, and we have found that yeast 

 washings, containing invertase, have no action on phaseolunatin. In these 

 two respects, therefore, invertase differs from the a-enzyme of Phaseolus 

 lunatus. 



The identity of the a-enzyme of Phaseolus lunatus with yeast maltase 

 cannot, however, be definitely established, since, although the range of 

 activity of the two enzymes appears to be similar, yeast maltase decomposes 

 a-methyl glucoside and maltose more rapidly than does the a-enzyme of 

 Phaseolus lunatus, and the latter hydrolyses phaseolunatin more quickly than 

 yeast maltase does. 



Since the enzyme preparation obtained from Phaseolus lunatus decomposes 



