TRAP. 



coak, or charred for fome diftance on each tide of the bafalt. 

 Thefe walls or dvkes of trap are fometimes twenty, thirty, 

 or even a hundred yards or more in thicknefs, and fometimes 

 range along through very estenfive dillrifts : indeed there is 

 reafon to beheve that many mountains of trap which are 

 arranged in one line along a coimtry, have been originally 

 dykes of enormous magnitude, through which vallies have 

 been excavated by diluvial currents, by partial fubCdence, 

 and by other agents which have in former ages changed the 

 furface of our prefent continents. See Veins, Mintral. 



Independently of theoretical opinions refpecting the 

 formation of trap-rocks, it is obvious that the malTes of un- 

 conformable overlying trap, and alfo the trap or whin-itone 

 in dykes, mull be of pofterior data to the rocks which 

 they cover or interfeft ; whereas the hornblende-rock or 

 primitive trap which is imbedded in gneifs, mica-Cate, or 

 clay-date, muft. be coeval with the rocks in which it occurs ; 

 and the fame is equally obvious with refpeft to trap-rocks 

 alternating with fecondary ftrata. Trap or bafalt fome- 

 times occurs in globular maffes compoled of concentric 

 fpheres, or of prifms of bafalt in a radiated diverging form. 

 The fpheroidal bafalt with concentric layers, frequently in- 

 clofes in its centre a mafs of bafalt fimilar to that which 

 envelopes it ; but fometimes the central mafs is of a harder 

 kind, or is a fragment of fome other ftone, as of Ihelly 

 lime-ftone. Many trap-rocks, in decompoCng, prefent a 

 fpheroidal ftructure which was not before viCble in the 

 mafs ; and it is obferred, that the balls of bafalt which fall 

 out or remain when the rock is decompofed, are harder than 

 the main body of the rock. This is the cafe with the bafalt 

 at Rowley in StaSbrdftiire, called Rowley Rag. See 

 Rowley Rag. 



The fpheroidal ftrufture is however not peculiar to 

 bafaltic rocks. Rocks of granite, in decompofing, fome- 

 times prefent globular maffes of harder granite imbedded in 

 the mafs. 



In columnar trap we find the columns of all fizes, from a 

 few inches to two hundred and fifty feet in length. The 

 form of the prifm or column is alfo various, and more or lefs 

 regular, having generally three, four, or five fides : the latter 

 or pentagonal form is the moft common. The diameter of 

 the larger columns is frequently four or five feet, pr more, 

 and fometimes is nearly equal to the height. 



Rocks of the trap-formation are alfo frequently obferved 

 to form conical mountains, or to occur as detached maffes 

 on caps on the fummits of other hills, and to occur at nearly 

 the fame level as reprefented Plate FV. Geology, fg. 3. c, d, 

 and e. In fuch inftances, it feems probable that the different 

 caps have originally formed parts of one continuous bed, 

 which has been broken either by the excavation of currents, 

 by fubfidence of the intermediate parts, or by the elevation 

 of the hills from their primeval pofition. 



The columnar ilructure and the overlying pofition fo 

 remarkable in trap-rocks, are aKo common to fome kinds of 

 porphyry, and to fienite, which are not mcluded by geo- 

 logiils among trap-rocks, but to which they bear a near 

 affinity. To take a juft and enlarged view of this clafs of 

 rocks, we muft leave the minute fubdivifions and claffifica- 

 tior.s of cabinet philofophers, and contemplate the natural 

 relaaons which thefe rocks have to each other, and the 

 gradations which may be traced between mountain maffes 

 poffefUng great difference in their external appearance. 

 Hornblende or augite forms an effential part of aknoft all 

 trap-rocks, as well as rocks of undoubted volcanic origin. 

 Now thefe minerals, though ckffed by mineralogiils as two 

 diftinft fpecies, are fo nearly alike in their conftituent parts, 

 thai we may coniider them as identicaJ, when they exiiled in 



a fluid Rate in the matter of the rocks of which they form 

 a part. Circumllances of minor import may eafily be con- 

 ceived to have changed the form of their cryftals, as we 

 know by experiment the flight caufes which occafioa 

 varieties of form in the cryftallization of the fame con- 

 ftituent ingredients in our laboratories. The conftituents of 

 hornblende and of augite, are files, alumine, lime, magnefia, 

 iron, manganefe, and a trace of potalh : the proportions of 

 each, which exift in the two minerals, do not vary more than 

 the proportions which exift in fubftances claffed by Haiiy 

 wkh hornblende, or between varieties of the fame fpecies in 

 many other minerals. 



We may regard then, in a geological view, the confti- 

 tuents of hornblende and augite as forming the fame 

 mineral, under whatever cryftalline forms they may prefent 

 therafelves in rocks. It is this fubftance which forms the 

 connecting link between granite, trap-rocks, and lava. 

 Hornblende enters as a conftituent part into many rocks of 

 granite. According to Patrin, maffes and veins of granite 

 occur in hornblende -rocks, and maffes and veins of horn- 

 blende in granite-rocks, in the Altaian mountains. Horn- 

 blende is found in fome of the granites of the Higher Alps : 

 it exifts abundantly in moft rocks of gneifs, which is only a 

 laminated granite. When the hornblende in granite increafes 

 in quantity, fo as to form a confiderable part of the mafs, 

 the rock is denominated fienite. When the hornblende 

 gives its colour to the rock, it forms greenftone. \\Tien the 

 parts are more intimately mixed, it becomes a compaft trap 

 or bafalt ; and bafalt and compaft lava bear in many in- 

 ftances fo near a refemblance, as not to be diftinguithed. 

 According to the account of fir George Mackenzie, beds 

 and even columns of bafalt frequently occur in Iceland, the 

 lower parts of which are a porous and vitreous lava or flag. 

 The clofe refemblance between the compaft lavas, and the 

 columnar bafalt in the vicinity of Etna, and other aftive 

 volcanoes, is fo great, that we believe it is extremely diffi- 

 cult, if not impoffible, to diftinguifh them from each other 

 by any well-defined charafters. The currents of lava 

 which have recently flowed, differ indeed from beds of 

 bafalt or trap by their porofity ; but Cordier has well ob- . 

 ferved, in the memoir before alluded to, that it is only thej 

 fuperficial part of modem currents of lava that are ex^ 

 pofed to our obfer\ation ; and to judge of their real natur 

 from the external fcorise, would be like judging of 

 contents of a vat of wine, from the froth floating on it 

 furface. The ancient currents of lava are fome of them or 

 immenfe fize, and have frequently been excavated or broken^l 

 by which the internal parts are difcovered, and thefe, as 

 before remarked, bear a near refemblance to many of the 

 regular rock-formations, and particularly to varieties of trap- 

 rock. 



Such being the leading charafters in the compoGtion, 

 flrufture, and pofition of trap-rocks, we need not be fur- 

 prifed that the earlier geologifts were difpofed to clafs them 

 all with volcanic prodnfts, and to affert that they had been 

 formed by eruptions iroia ancient volcanoes, in a manner 

 fim ilar to the formation of modern currents of lava. This 

 hypotheCs, advanced before manv phenomena attending vol- 

 canic eruptions were known, and when the effefts of heat 

 combined with preffure had not been inveftigatcd, was ex- 

 pofed to many weighty objeftions. Another theory, which 

 afcribed the formation of trap-rocks to aqueous depofitions, 

 was warmly fupported by many geologifts on the con- 

 tinent, and particularly by M. Werner. He fuppofes, that 

 after all the primitive and cojiformsbie rocks had been 

 depofited, the primeval waters which covered the higheft 

 mountains funk flowly to their prefent level. 



4 From 



