256 COSMOS. 



action it produces, must therefore compose the third class of 

 the fundamental forms of rock. 



Endogenous or erupted rocks (granite, porphyry, and mela- 

 phyre,) produce, as I have already frequently remarked, not 

 only dynamical, shaking, upheaving actions, either vertically 

 or laterally displacing the strata, but they also occasion 

 changes in their chemical composition, as well as in the nature 

 of their internal structure. New rocks being thus formed, 

 as gneiss, mica slate, and granular limestone (Carrara and Pa- 

 rian marble.) The old silurian or devonian transition schists, 

 the belemnitic limestone of Tarantaise, and the dull grey 

 calcareous sandstone (Macigno} which contains alga? found 

 in the northern Apennines, often assume a new and more 

 brilliant appearance after their metamorphosis, which renders 

 it difficult to recognise them. The theory of metamorphism 

 was not established until the individual phases of the change 

 were followed step by step, and direct chemical experiments 

 on the difference in the fusion point, in the pressure and time 

 of cooling, were brought in aid of mere inductive conclusions. 

 Where the study of chemical combinations is regulated by 

 leading ideas,-' 4 it may be the means of throwing a clear light 

 on the wide field of geognosy, and over the vast laboratory of 

 nature in which rocks are continually being formed and 

 modified by the agency of subterranean forces. The philo- 

 sophical enquirer will escape the deception of apparent 

 analogies, and the danger of being led astray by a narrow 

 view of natural phenomena, if he constantly bear in view the 

 complicated conditions which may, by the intensity of their 

 force, have modified the counteracting effect of those indi- 

 vidual substances, whose nature is better known to us. Simple 

 bodies have, no doubt, at all periods, obeyed the same laws of 

 attraction, and wherever apparent contradictions present them- 

 selves, I am confident that chemistry will in most cases be 

 able to trace the cause to some corresponding error in the 

 experiment. 



* See the admirable researches of Mitscherlich, in the AWiandl. der 

 fieri Akad. for the years 1822 and 1823, s. 25-41 ; and in Poggend. 

 Annalen, bd. x. s. 137-152, bd. xi. s. 323-332, bd. xli. s. 213-216. 

 (Gustav Rose, Ucber Bildung dcs Kalkspatlis und Aragonits, in Pog- 

 gend. Annalen., bd. xlii. s. 353-366 ; Haidinger, in the Transactions of 

 the Royal Society of Edinburgh, 1827, p. 148.) 



