TRANSACTIONS OF THE SECTIONS. 49 
that this smoke acted chemically, by preventing the oxidation of the iron; and thus 
the change from the dark colour of the clay to red, which usually occurs in burning 
pottery and bricks, was prevented. 
᾿ς Reference was then made to a medallion of the pavement representing Flora, in 
the first drawing of which the head-dress and flowers held in the hand were coloured 
verdigris-green, the hue these objects presented on being exhumed ; but as this was 
unsatisfactory in chromatic arrangement, the author suspected some subsequent che- 
mical change; and on scraping away the green from the surfaces of the tesselle in 
question, a beautiful ruby glass presented itself. New drawings (which were also 
exhibited) were then made, with ruby instead of green colour; the result of which 
was, that what was before inharmonious in colour and grouping, at once assumed 
harmony in these respects, and became perfectly intelligible. An analysis of the 
glass, made by Professor Voelcker, showed the cause of change from ruby to green 
to have been due to the fact, that the antique ruby glass had derived its colour from 
peroxide of copper, and that the tessellze had become covered with carbonate of copper 
from a decomposition of their surfaces. 
On the Action of the Soap-test upon Water containing a Salt of Magnesia 
only, and likewise upon Water containing a Salt of Magnesia and 
a Salt of Lime. By Ducatp CampseLt, 3.0.5. 
Experiments were first made upon water containing a salt of magnesia alone in 
different proportions, in order to ascertain if the soap-test acted throughout with mag- 
nesian solutions as it does with lime. For this purpose, 19-2 grains pure dry sulphate 
of magnesia (MgO, SO3), the equivalent quantity of that salt to 16 grains carbonate 
of lime, were dissolved in a gallon of distilled water. This solution was considered 
as a standard of magnesia of 16° hardness. Fifteen other standards were prepared 
from this by proportional dilution with distilled water, in the same way as the stan- 
dards of lime are recommended to be prepared by Professor Clark in the specification 
of a patent printed in the ‘ Repertory of Patent Inventions’ for 1841. 
The action of the soap-test in producing perfect lathers with these standards coin- 
cides, or nearly so, with the action of the soap-test upon like standards of lime up to 
the sixth degree ; but from that point the'magnesian standards begin not to require 
so much soap-test as the lime; and as the standards increase, this difference in soap- 
test increases till the magnesian standard of 16° requires only 19᾽6 soap-tests, whilst 
the lime standard of 16° requires 32. 
Standard solutions were prepared containing 16°, 12°, 8°, 6°, 4°, 2° lime in a gallon 
plus 1°, 2°, 3°, 4° magnesia, and so on up to 16°. 
Less soap-test is requisite to cause a perfect lather in most of these solutions than 
is requisite for the standard of lime alone contained in them ; or, in other words, 
the magnesia appears to soften the lime standards, and this peculiarity increases as 
the magnesia increases; for example, a standard of lime of 16° takes 32 test measures ; 
a standard of lime of 16°-+- 1° magnesia takes 51:6, and a standard of lime of 16°-4+-16° 
magnesia, 27:9. 
As the standards of lime decrease, the action of a few of tle lower degrees of mag- 
nesia in softening is modified, till solutions of the standard of lime of 2°, plus the first 
4° of magnesia, take nearly as many soap-test measures as if they were entirely lime 
solutions. 
In a number of these standard solutions, when a perfect lather had been produced 
by a minimum quantity of soap-test, the addition of a small quantity more reduced 
the lather to a curd; great agitation produced the lather again, and stronger than it 
was before the last addition of soap-test. If beyond a certain amount of soap-test 
however had been added, a new action showed itself, which was that the lather cannot 
be restored properly again until a certain amount of soap-test had been added. The 
additional quantities of soap-test were in every case noted when this peculiarity was 
observed. 
The inferences drawn from the experiments are as follows:—1. That water con- 
taining sulphate of magnesia alone acts towards the soap-test in producing with it a 
perfect lather, similarly or nearly so, as does water containing a lime salt alone, but 
1850. E 
