38 MR. JOHN GRAHAM ON CONSUMPTION 
sawdust, spent madder, and logwood chips, with various 
degrees of advantage. Muriaticacid proved a most effec- 
tive agent, indeed somewhat too much so. Although I 
never added more than two-thirds of the quantity of that 
acid necessary to saturate the lime known to be in the water 
as carbonate, yet the action continued not only to the point 
of the disappearance of the remaining third of the carbonate 
of lime, but went further. It had an action, also, on the 
oxide of iron surrounding the rivets, and existing between 
the edges of the plates; so that after six weeks, the usual 
duration of an experiment, it was found that the iron plates 
were not ouly clean and bright like silver, but also that 
the joints had been acted on to a considerable extent. 
The rust did not appear to be dissolved, but only loosened, 
nor did the metal of the plates appear to be corroded, 
for water from the boiler tested daily showed no iron. 
The water used in these experiments was from the Lan- 
cashire “Red Rock,” hard in quality, containing both 
sulphate and carbonate of lime, with a little iron. It re- 
quired forty measures of Dr. Clarke’s soap test liquor to 
soften it. The scale, when allowed to form, was com- 
posed in 100 parts of 22.5 carbonate of lime, and 77.5 
sulphate of lime. 
Two facts were noticed as regards the tendency of hard 
water to scale. First, that the sulphate of lime separates 
from the water only when in contact with solid objects, 
such as the bottom and sides of the boiler, or solid matters 
floating in the boiler, such as sawdust and logwood chips 
(hence their use); and that no precipitation takes place until 
the water has been concentrated by continued evaporation 
down to the condition of a saturated solution of sulphate 
of lime, or to that point which may be termed the “ salting 
point.” Secondly, carbonate of lime and mud are princi- 
pally deposited in the midst of the water, and have but little 
