ON THE ACTION OF AIR AND WATER UPON IRON. 233 
The iron must be, for the greater part, in the state of mag- 
netic pyrites (6 FeS+Fe S,), and the remainder in the state of 
persulphuret. The gray paste of the crust is coloured like- 
wise by the sulphuret of iron [Berthier does not say which]. 
*«The production of sulphuret of iron at the surface of this 
anchor in the Seine in contact with wood, is not surprising, 
since we know that its waters contain sulphate of lime, and 
that sulphates in solution are converted slowly into sulphurets 
in presence of organic matter, with production of carbonate of 
lime. 
“The carbonic acid, dissolved in the water, must have had 
likewise some influence on the results of this reduction, and 
it is probably to its presence that we may attribute the forma- 
tion of the persulphuret of iron.” 
These facts sufficiently indicate the accordance to nature of 
the theory of reaction of putrid waters which has been advanced, 
and are fertile in conclusions of practical importance to the en- 
gineer occupied in works of construction in iron, connected with 
the docks or harbours of large towns, or in similar situations. 
173. The average minimum corrosion is indicated (by Tables 
V. and VI.) as occurring in clear river water, containing air in 
combination. This difference below the index of corrosion for 
clear sea water is to be accounted for upon the general princi- 
ples already laid down*; and as subsidiary causes influencing 
the results are to be noticed, the absence of all extraneous cor- 
rosive agents, and the circumstance that the coat of oxide of 
iron formed in fresh water adheres obstinately to the iron, often 
forming imperfect crystals, of brown hematite (fer oligiste), 
and is not removed in a loose pulverulent form with the same 
€ase as it is in sea water, and hence acts in some degree as a 
cloak, or partially impervious covering, to defend portions of 
the metal from further action. 
- 174. It will be further observed, from Tables III. and VI., 
that wrought iron suffers greater loss by corrosion in hot sea 
“water (temp. 115° Fahr.) than under any other circumstances— 
an important result as respects the construction of marine en- 
gines and boilers. Upon the latter point a special set of ex- 
periments are in progress, having in view principally the points 
of inquiry suggested +; and should the results of these experi- 
ments give a decidedly advantageous point of concentration at 
which to work marine boilers (other circumstances being consi- 
| dered), there will be no difficulty, by the aid of the brine-pumps 
now occasionally used in first class steamers, and of Mr. Sea- 
* §§ 167, 168, 169. + First Report, § 78. 
