as illustrative of Geology. 71 



played in such a way, that we cannot suppose it to exist in 

 another state immediately previous to the crystalline one. It is 

 well known that carbon occurs in various ways in the grey crude 

 iron, namely, partly in chemical combination with the iron, partly 

 as graphite mixed with it. That the quantity of carbon which 

 is contained in the liquid crude iron and which is not able to 

 combine the iron chemically, in the moment of rigidity, and 

 even partly while the iron is still fluid, is dissipated in vapour, 

 and passes from this state immediately into the crystalline, ap- 

 pears susceptible of proof from the circumstance that the gra- 

 phite is observed not only in the interior, but even on the actual 

 surface of the crude iron, like crystals of snow on ice, formed 

 by the ascending vapour, — that its crystals are met with on the 

 slags, which cover the iron in the furnace, and even in their in- 

 ternal cavities, — and that the graphite is sometimes found even in 

 the interior of the wall, and not unfrequently associated with 

 those bodies (such as silica), whose penetration in a state of va- 

 pour we are compelled likewise to assume. The more slowly 

 the cooling proceeds, the more unconfined the space, and the 

 smaller the external pressure, the more distinct is the crystal- 

 line structure of the graphite. The largest crystals are found 

 in the cavities of the crude iron on its actual surface, and particu- 

 larly on the slags. The more slowly the crude iron cools, the more 

 distinct appear its scales ; whereas, in cases of rapid cooling, for 

 instance, when crude iron is plunged in water, the scales are in- 

 distinct. In grey crude iron the graphite is most usually more or 

 less uniformly mixed with the iron : in many varieties, however, 

 among others in the Swedish and Norwegian crude iron, the mix- 

 ture is not uniform ; in which case, the graphite sometimes forms 

 thecompletion of globular spaces, and then its crystals are grouped 

 concentrically, as we observe in amygdaloids, resembling, for 

 instance, the appearance of the scaly chlorites. This phenome- 

 non might confirm the assumption, that graphite vapour formed 

 single bubbles in crude iron ; and this leads further to the sup- 

 position, that the cavities in rigid crude iron in which graphite 

 crystals are often found, are partially generated from the carbon 

 vapour after it has effected its escape. Particular notice is due 

 to the appearance of graphite in the cavities of a slag of a blast- 

 furnace at the same time with iron, which in a botryoidal form, 



