386 Hale — Initiative Action of Iodine and other 



end reactions in blank. It will be noticed that the plain iodine 

 reading is not sharp unless potassium iodide be present, and 

 this is true of the starch solutions. The ordinary starch solu- 

 tion mentioned in this table and in the following table was a 

 different one from that previously used. It was freshly pre- 

 pared and hence shows a less amount of the impurity present. 

 The burette used took four to five drops to make 0*l cm3 , or 

 approximately one drop equaled 0-022 cm3 . ' 



It will be seen that the potassium iodide starch solution gave 

 the sharpest and best end reaction, that amorphous amylodex- 

 trin lost one drop of iodine, and amidulin and plain iodine 

 two drops. However, the presence of a crystal of potassium 

 iodide rendered them all sharp to a drop, with the exception 

 of amidulin, which explains the closeness of the results in the 

 next table. In this their delicacy was tested in regular titra- 

 tion with a fresh arsenite solution, and with a slightly stronger 

 iodine solution. 



Table V. 

 Volume 125 cm3 . 



As 2 3 sol. n/10 1 sol. Starch sol. Color. 



.3 



cm°. cm°. cm 



50 



49*38 1 KI starch permanent purplish 



49*40 good blue 



49-40 1 " (4 crystals KI) good blue 



49 40 2 ordinary slow-fading purplish 



49*42 permanent blue 



49*38 1*5 amorphous amylo- permanent purplish 



49*40 dextrin deep blue 



49*42 5 amidulin permanent purplish 



49*44 deeper purple 



49*55 25 ordinary abundant red 



While these results are very close and the discrepancies 

 might be ascribed to manipulation, yet the loss on 25 cm3 of 

 ordinary impure starch freshly made, can hardly be over- 

 looked. Moreover, these titrations were made at a slight dilu- 

 tion, and the slight loss hinted at in the case of amidulin is in 

 keeping with later experiments. The introduction of con- 

 siderable potassium iodide made no difference in the sharpness 

 of the potassium iodide starch, and since this preparation gave 

 the best blue color it was used in a series of parallel titrations 

 in which readings were made with the starch blue and then 

 with plain iodine alternately in order to eliminate accidental 

 errors. Starch was added subsequently to corroborate the 

 plain iodine reading, as the yellow is very delicate. The abso- 



