174 ROYAL SOCIETY OF CANADA 



for the various tests upon the same solution did not always agree, prob- 

 ably due lo ihe addiiion of a little too much or too little of one or the 

 other re-agent. As the colour of the test changes with age, fresh tests 

 from the known solutions had to be taken from time to time. 



The Bach test referred to, while much more convenient to apply, 

 is not as sensitive and did not give as good quantitative results as the 

 starch KI test. The colour from the strong solutions was hardly any 

 deeper than that from much weaker ones. The stronger the solution 

 the quicker the colour appeared, but it also faded quicker. About 1 part 

 in 200,000 was the limit of this test. The titanium sulphate test was 

 tried, but failed altogether, as 1 in 40,000 was its limit. 



From the results of the starch KI test tlie greatest yield with the 

 aluminum was judged to be between 1 in 50,000 and 1 in 100,000, while 

 that from the zinc was between 1 in 50,000 and 1 in 70,000. 

 Magnesium gave not more than 1 in 300,000. 

 In reviewing these results we see that the metals whose oxide or 

 hydroxide is not protecting decompose the peroxide rapidly, while those 

 whose oxide or hydroxide is more adherent do so very much more slowly. 

 When air is bubbled through the water, the formation takes place in 

 spite of the counter reaction of decomposition, but decomposition takes 

 place when the air current is not kept up. 



It seems probable that the yield of H^O^ is not always checked by 

 the fouling of the metal, but that at some point the formation and 

 decomposition balance. This is borne out by the fact that in some of 

 the experiments with aluminum the yield was the same, although there 

 was more metal surface in one than in the other. If too little surface 

 is supplied the action is checked by the fouling of the surface. Owing 

 to the form of the metals used, turnings, scrap, etc., it was impossible 

 to estimate the surface exposed. 



Whether in the case of iron the peroxide is formed and 

 immediately decomposed or not formed at all, is a question on which 

 there is considerable difference of opinion, but as far as the writer is 

 aware, it has never been detected. 



