360 



Prof. H, E. Armstrong and Mr. E. Horton. [Mar. 3, 



On reference to E. F. Armstrong's account of his work* it will be found 

 that his original observations are, in most cases, affected by similar " errors." 

 It would seem that the changes taking place in the solutions examined are 

 not only those which are involved in the alteration of the glucose originally 

 produced into the equilibrated mixture of isomerides — in fact, the substances 

 present in the solution of the enzyme are also affected by the alkali. 



Our experiments have afforded abundant evidence that such is the case, as 

 will be pointed out later on. Moreover, we find that the Phaseolus enzyme 

 is itself sensitive to alkali ; thus a solution of which the initial rotatory power 

 was + 0°'01 was found to be considerably altered, the rotatory power changing 

 from — 0°'02, the value after adding alkali, to — 0°'9o at the end of four hours. 



Under these circumstances we have deemed it desirable to repeat some of 

 E. F. Armstrong's experiments, in order to verify his statements — especially 

 as the instrumental appliances at our disposal are far better than those over 

 which he had command. The following series of observations made with 

 methyl-/S-giucoside and emulsin are a complete justification of his method : — 



Time of 

 action. 



Initial rotation. 



Rotation after 

 addition of 

 ammonia. 



Change in 

 rotation. 



hours. 



o 



o 



c 



0 



-2-83 



-2-80 



+ 0-03 



1 



-2-30 



-2-14 



+ 0-16 



1 -5 



-2-03 



-1 -85 



+ 0-18 



2 



-1 -74 



-1 -59 



+ 0-15 



2-5 



-1 -53 



-1 -38 



+ 0-15 



4 



-0-71 



-0-61 



+ 0-10 



4-5 



-0-54 



-0 -44 



+ 0-10 



6 



-0-12 



-0-04 



+ 0-08 



24 



+ 2-63 



+ 2-62 



-0-01 



Attempts to carry out a similar experiment with maltose were unsuccessful 

 owing to our failure to obtain enzyme extracts which remained clear on the 

 addition of alkali. It may be mentioned that in E. F. Armstrong's experi- 

 ments the liquid was filtered to remove tlie turbidity occasioned by alkali ; 

 we have suuglit to avoid any such operation and have added the alkali 

 directly to the liquid in the tube. 



AnM/5 phaseoluuatin solution containing 25 c.c. of phasoolunatase solution 

 per 100 c.c. was kept in a Jena flask at 37°. At hourly intervals, 20-c.c. 

 samples were witlnhuwu, inix(^d with 5 c.c. of alumina cream, rapidly iiltered 

 through asbestos and exaniiniMl polarimetrically at 2.")° in a 2-decimetre tube. 

 Then two drops of soda solution wore added and the rotation again determined. 



♦ *Chein. Soc. Trans.,' 1903, vol. 83, p. 1305. 



