234 Mr. T. M. Rpade on the Effects of Contraction 



Westmoreland, to find so little evidence of contraction. If, 

 PS we have every reason to believe, the granite has once been 

 in a state of igneous fluidity, it would at first sight seem 

 reasonable to expect more evidences of contraction than what 

 are usually manifest. The late David Forbes was so much 

 struck with this in Norway that he instituted a set of experi- 

 ments to test the statement of Bischof that granite contracts 

 from 10 to 25 per cent, in volume in cooling. His results 

 in the casting of highly siliceous slags showed only a 

 contraction of from 0'()14 to 0*028. Forbes did not attach 

 great weight to these experiments as accurate determinations 

 in consequence of the difficulty of getting a perfectly solid 

 casting, but considered that they proved Bischof to have much 

 overestimated the contraction of granite. Let us consider 

 what the process of contraction ^s, and in what way granite- 

 masses are likely to be affected by it. 



In casting ironwork it is usual to allow in the pattern ^^ of 

 an inch per linear foot for the contraction of the iron=j^Q of 

 the whole length, or a contraction in volume of ^q. This 

 is not, however, the contraction from a fluid state, for the 

 exterior cools first, and, at the time of solidification, is as large 

 as the pattern. The contraction of iron in passing from the 

 fluid to the solid state or in " freezing '^ is therefore not de- 

 termined by this practice, and I am not av\are of any reliable 

 determination of its amount ; indeed some have contended 

 that it expands in freezing, because cold iron will float in 

 molten iron. 



Be this as it may, I think we can get some hints of the 

 operations of nature from the processes of the ironfounder. 

 For instance, in casting a large sphere, say 2 feet in 

 diameter, it is necessary, to prevent the honeycombing of 

 the central part, to keep feeding the ball as the metal cools. 

 It would seem that the outer skin takes the exact size and 

 form of the mould on solidification, and the metal cools and 

 congeals from the exterior to the centre. By keeping the 

 centre fed with new molten metal a sound casting is 

 insured. In the same way any one who has been in the 

 habit of making castings in lead is aware of the sudden 

 disappearance of the lead in the feeding-throat on congelation 

 taking place, so that unless a good plug be left a deep hole 

 will occur in the casting at the point where the metal has 

 been poured in. Both these phenomena point towards 

 sudden contraction on congelation. 



Let us now consider the case of a granite-mass. It is 

 extremely probable that in many cases these masses are more 

 or less like what Mr. Gilbert has called Laccolites or stone 



