ACTION OF WATER ON IRON. 283 



his structures, and enabling us, by analysis of the least and 

 most corrodible, to see on what these properties depend. Hot 

 and cold blast iron are included in this box, and also specinaens 

 of the same cast iron chilled, and cast in loam, dry sand and 

 green sand, cooled rapidly and slowly, and some with protected 

 surfaces, (not electro-chemically protected however.) as will be 

 again alluded to. This box was sunk and moored in Kings- 

 town harbour in 3\ fathom water, at half tide, at the second 

 mooring buoy in from the western pier head, at one o'clock, on 

 the 3rd of August (1838), on a bottom of clean sharp sand; 

 temperature of the water, 58° Fahr. It is proposed being 

 weighed and examined once every six months if possible, and 

 at the expiration of a year the specimens weighed accurately 

 and again put down, and so weighed year by year for at least 

 four years. By this, not only the actual amount, but the rate 

 of progress of corrosion on every specimen will be determined, 



71. The object of casting the inch-square bars 12 in. long, of 

 each sort of iron, at the same time with those exposed to the 

 sea and fresh water, is in order to have specimens, comparable 

 in all respects with these as to constitution and texture, whose 

 specific gravities have been, or are in progress of being, taken, 

 and whose chemical or physical properties can be in future de- 

 termined, should the progress of the experiments on the exposed 

 pieces render such desirable. 



There is the utmost variability of structure and composition 

 amongst these specimens, as will be observed by referring to 

 the column of observations in the table, but can only be fully 

 perceived by inspection of the castings, when fresh broken. 

 As an accurate knowledge of the specific gravities of these spe- 

 cimens was of some importance in several respects, and chiefly 

 as being a check upon the weighings, and upon the soundness 

 and dimension of the rectangular pieces of iron, submitted to 

 experiment, some pains were taken in arriving at the best and 

 most expeditious mode of proceeding. 



The common mode of taking the specific gravity of solids by 

 weighing in air, and then suspending from a silk fibre or hair in 

 water, is subject to many inconveniences and sources of error, 

 as the variable quantity of the suspending thread wetted and 

 immersed, its capillarity and the resistance of the fluid to free 

 vibration of the beam, &c. &c., a modification was therefore 

 adopted of the plan of weighing the solid in air and immersed 

 in a given volume of water, which is weighed along with it. 



But as the cast irons, if broken into fragments sufficiently 

 small to go into a common specific-gravity bottle, would be 

 likely to involve air bubbles hard to be extricated, each speci- 



