376 Mr. T. Andrews. Action of Tidal Sfreams on [June 18, 



metal in each case, from which it is manifest that a diffusing mixture 

 of salt and fresh water, although containing a comparatively small 

 proportion of saline matter, is capable of setting up an amount of 

 corrosive mischief far greater than that arising from the action of 

 sea water only, the increased loss varying from about 15 up to 50 per 

 cent., according to the nature of the metals. From a reference to 

 column 3 it will also be seen that the combined action of the sea 

 and distilled water (although showing only such a slight saline com- 

 position and difference of salinity as indicated on Table B of analysis) 

 whilst diffusing, produced an amount of local galvanic action in the 

 same plate 0, sufficient to induce therein much greater corrosion than 

 in the case of plate B, although the latter was constantly immersed in 

 the sea water (containing 2276*8 grains per gallon of sodium chloride), 

 and was at the same time metal positive, as indicated by the galvano- 

 meter. 



Table C. — Loss by Corrosion of Metals in Sea Water, and of the 

 same Metal exposed to the Galvanic Action of diffusing Salt and 

 Fresh Water during Fifty- two Weeks. Results in grains per 

 square foot of surface exposed. 



Description of metals. 



Common cast metal 

 (No. 2) 



" Soft " Siemens-Mar- 

 tin steel 



"Wrought iron 



" Soft " cast steel. . . . 



"'Soft" Bessemersteel 



Best cast metal (No. 1) 



" Hard " Bessemer . . 



0-670 



0-170 



None 

 0-460 

 0-150 

 0-390 

 0-510 



Col. 1. 



o3 



^2 



CD *- 



Tfl 



235-42 



259 -94 



270-28 

 278 -54 

 300-26 



308 -68 



309 -42 



Col. 2. 



rd m 



.2 ^ 



3,3 



i> .9 

 o3 pq 



240-07 



262 -38 



284-16 

 343 -98 

 278 -92 

 326 -99 

 304 -39 



Col. 3. 



Pi 



CO 



8 3 I* 



289 -03 



395-52 



365 -98 

 349 -29 

 345-16 

 354-39 

 402 -31 



Col. 4. 



236-75 



344 -34 



310-60 

 249 -38 

 252 -92 

 269-61 

 309 -27 



The experiments recorded in this and the former paper indicate 

 therefore, that the tidal action on any vessel or metallic structure of 

 sea and fresh water whilst diffusing, is (in the case even of the same 



