194 
ME. EOBEET MALLET ON VOLCANIC ENEEGY. 
Measurements of the wood “ pattern,” “ often of some feet in length,” and of the 
solidified counterpart in fused rock cast from it, according to Messrs. Chance’s statement, 
proved them to he precisely the same size, from which it is inferred that there was no 
contraction at all in the rock by cooling. 
Now this is exactly the same mistake that has been made by Hr. Percy in his ‘ Metal- 
lurgy of Iron,’ where he infers the coefficient of total contraction of cast iron from the 
difference between the size of the wood pattern from which the casting is made and the 
casting itself. 
In both these cases the pattern, in order to be extracted from the damp “rammed up” 
sand, has to be “ rapped ” by the workman — that is to say, he strikes the pattern in all 
directions so as to make the cavity in the sand bigger than the size of the pattern, the 
difference being for clearance to enable the withdrawal of the pattern. If the sand 
mould be then dried and heated to something like redness, it will by expansion (on the 
whole, for there may be some little contraction at first on drying) cause the cavity of the 
mould to become still larger, and it is at this temperature that it receives the liquid 
rock. 
Hence the equality in size between the wood pattern and the again solidified rock, 
even if exactly the fact, does not prove that there was no contraction, but that the 
amount of contraction was equal to the enlargement of the cavity of the mould, equal 
to the bulk of the pattern by the “ clearance ” given to the latter by “ rapping ” and by 
the further enlargement of the mould itself by heating ; and as neither the volume of 
the clearance (which is never the same in any two cases) nor the enlargement by expan - 
sion (which may vary with the quality of the materials of the mould at different times, 
as well as with the temperature) are known, so no conclusion whatever can be drawn 
from those data as quoted from Messrs. Chance. 
138. Then as to Mr. Forbes’s own results, we surely cannot rely upon measurements 
made upon blocks of slag cast into iron ingot moulds of only 10 inches long, 6 inches 
wide, and 6 inches deep = 360 cubic inches. 
The contractions stated to be thus deduced at Eidfors Iron Works, Norway, of the 
slag (blast-furnace no doubt), which were from 1^ to 3 per cent, (wide limits, the largest of 
which must yet be within the errors of these experiments), cannot be relied upon when 
we are left in the dark as to the temperature of the moulds, and therefore their actual 
size at the moment of consolidation, or whether any and what final cavities formed inside 
the solidified mass, and take into account the impossibility of measuring with any great 
exactness the true dimensions of such a small and even not rectangular block of slag with 
necessarily always a more or less rough and uneven surface. 
139. Mr. Forbes’s further experiments on basic slags from Staffordshire blast-fur- 
naces, “ cast into sand moulds,” are vitiated through just the same reasons as those on 
the Rowley Rag. 
140. Lastly, Mr. Forbes adduces some measurements made at Birmingham on blocks 
■of cast glass cast in iron moulds, which either showed almost no contraction or (according 
