184 THE ROYAL SOCIETY OF CANADA 



in potential between one another in the presence of an electrolyte. 

 Careful research has produced a list of the metals in the order of their 

 potential. In our experiments we considered the following only, 

 calcium, magnesium, aluminium, zinc, iron, lead, hydrogen and 

 copper. These are named in the order of ascending potential and in 

 this order is their power to displace one another. Thus in the presence 

 of water even a single '* commercial " metal will set up '' local currents " 

 which are active agents in corrosion. 



The foregoing formed the chief consideration which governed 

 the plan of the enquiry, but there were several other features, such as 

 abrasion of the metals, differences in temperatures at various points 

 of the metal, points of contact of the metal with containing vessels, 

 &c., &c., that were found necessary to take into consideration. 



It may be mentioned incidentally that complete analyses of 

 raw and treated waters were made, and it might be interesting to 

 note that, when the treatment is skilfully and carefully carried out, 

 the alumina content of the filtered water is practically identical with 

 that of the raw water. It was also found that "mechanical filtration" 

 removed approximately 60 per cent, of the dissolved organic matter. 



In view of interesting figures respecting corrosion later obtained, 

 it may be mentioned here that magnesia (MgO) was experimented 

 with as a coagulent for peaty waters. This compound was found 

 particularly effective as a decolorizing and clarifying agent, yielding 

 a perfectly colorless and clear water after a few hours sedimentation. 



From what has already been said relative to the electrolytic 

 theorj'- of corrosion, the importance of the measurements of electrical 

 conductivity might be surmized. However, this factor becomes 

 so seriously involved when comparing waters of different character, 

 from the fact that the various mineral salts contained therein are 

 variable conductors of electricity and yet in themselves may be even 

 anti-corrosive, that it is very difficult to draw definite conclusions 

 as to the corrosivity of a water from its conductivity alone. Read- 

 ings of this nature gave figures in the following ratio — ^treated 125, 

 raw 87, purest water to be obtained 2 to 3. From these figures it 

 is seen that " mechanical filtration " increases conductivity from 25 

 to 50 per cent. 



Methods of Experimentation. 



In carrying out these experiments care was taken to make con- 

 ditions in all cases uniform, so that the results, though perhaps not 

 absolute, might be at least comparative. The metals used were em- 

 ployed in the form of wire. These were cleaned and polished and cut 

 in equal lengths of 18 inches. They were accurately weighed, and 



