
ON THE ACTION OF AIR AND WATER UPON IRON. 7 
a non-homogeneous surface, and aiding thus in the transfer of the oxides 
formed and not dissolved. 
298. On opening the several boxes of cast and wrought iron after immer- 
sion, the special appearances of each specimen were as before marked, and the 
form of corrosion is stated in the respective tables. The phenomena gene- 
rally were much the same as in the prior exposure, with the exceptions already 
adverted to. All the cast iron pieces in sea water were irregularly covered 
with a thin coat of carbonate of lime. . 
299. It may be observed, that the decrement of the rate of corrosion of all 
the cast iron specimens is considerable in the second immersion. For the 
purpose of obtaining the amount of corrosion in the first immersion, the coat 
of plumbago and rust formed was necessarily removed, and this, contrary to 
what might have been presumed, I am now enabled to prove is the chief cause 
of the decrement. 
300. Six equal parallelopipeds of the same bright gray cast iron with planed 
surfaces were immersed in separate vessels in sea water, slightly acidulated 
with muriatic acid, and frequently renewed. Each of the six was removed 
at successive intervals of thirty days; the coat of plumbago and rust removed 
from the piece, which was dried and weighed. The following table gives the 
results. c 
The weight of the original pieces was in every case = 1060 grains. The 
temperature of the menstruum 54° to 67° Fahr. 
TABLE B. 

Weight of each Losses of 
: Specimen after Me Weight by Weight divided 
exposure for the} Corrosion. |Py the times of 
No. of |No. of days’) Original Weigh 
Speci- | exposure to} of each Specimen | 
“men. | Corrosion. | before exposure. 

above times. exposure. 
Grains. 
1 30 1060 1057°2 2°8 2:8 
2 60 1060 1054°6 54: 2:7 
3 90 1060 1051°2 88 2:9 
4 120 1060 1048°1 11:9 3°0 
5 150 1069 1045°9 14+] 28 
6 180 1060 1041 +4: 186 31 
301. The specimens were all cast from the same mass, chosen with special 
regard to the uniformity of its texture ; and the results of the preceding table 
show, that when the coat of plumbago and rust formed remains untouched 
during the whole period of immersion, the amount of this, or the actual 
‘loss of metal, is very nearly in proportion to the time of reaction, showing 
that the coating of plumbago and peroxide is negative with respect to the 
metal and aids in its corrosion. This result however applies much more to 
Corrosion in salt than in fresh water, wherein the coat of rust formed is much 
harder and less porous; and hence, although still negative to the metal, par- 
tially defends it mechanically from corrosion. 
802. It may be noticed, that on taking up the kyanized oak-boxes of spe- 
cimens from Kingstown Harbour, after two years’ exposure, the timber was 
found perforated nearly through a thickness of two inches by the Limnoria 
terebrans, whose ravages are thus proved not to be arrested by kyanizing. 
303. I now proceed to some notice of the series of experiments in Table 
IX., on iron exposed to the atmosphere, &c. at Dublin. The forms in which 
cast iron corrodes in water df various sorts have been heretofore minutely 
