period in which their ingredients retain a viscid consistency p will , from the expansion eon- 

 sequent nix cryttnJl iialio a , be ei posed Dot only to mutual la tern] pressure but to pressure 

 from beneath , and this will lend r according Id the tale of refrigeration and other circum- 

 ittneet # raws or leu to obscure or even oblLlercitu t lies spherical form. Where I here ii a 

 considerable proportion of mlm iho eorjeeiilrie laminar arrangement will if ill be preferred. 

 1 cannot follow out llm vie it here; but the experiment on the gradual opraling or molten 

 ruck first made by Mr, Gs&gokt Witt, and freqUEntly repeated since , would seem to c*- 

 plain »U thi: gradations or igneous rocks. With reference to the hypogene rocks., in 

 order 1o conceive how the dillermt Kpceici may In- produced from ijjneous fluid beneath the 

 earth's. CfOilj we need only consider under what different conditions the fluid must bate 

 eiisted as fa tranquillity t agitation, or motion. t th* eitcnE and Corm of refrigerating 

 HiirfiiceSjt «* lhe nature, thickness, and pressure of lite refriff era ling jn&m% IheniwTtci. 

 whether rock y or sea, or both, — the rate of diminution of temperature, — ond the prDpor- 

 lious of chemical ingredients. Many of these condt lions may hate \aried in different por- 

 tions of the same great fluid ma*^ pad ot different stores in the proce** of it* cryatallita- 

 lion and scilidltlcaliom The phenomena, attending Mr. Witt's experiment E\cn seem to fIicw 

 that the ([lobular, cubcidalj ur prismatic coMcrrfioHfir^ structure of micaceoui granite mi^ht 

 pass into the Laminar guei&sosc structure merely Ihroagh oscillations in temperature during 

 solid Miration An analogous passage From nodules into later*, and., in the latter, the 

 arrangement of the crystals of one of I wo or more d liferent mineral* in continuous parallel 

 laminae , charade ri»cs some Tolcauie rock*. 



If e spans] on on crystallisation he attributable to the nether hypogene fluid masse* 7 and 

 they ore gradually solidified from above downwards r a slow upheaval of ihe superincumbent 

 crust must attend their sol idi heat ion when tliey are of sufhVicnth \\reai tiltnL, This roay 

 Mic cause of 1 tie i-lc-vallon i>f Srftridiitftr.ia :iurj i.Llu-r > > hi e i I r i r ^ . :A pre* en I tn [HnjfJPGB*, 



Whatever may be the origin of the bedded structure of the hypogene rocks , it is to u 

 similar itralUicoUon , combined with the vertical joints, that the form* assumed by the ci- 

 ]>Osed masses on Pufo Hhin must probably be referred , and to these I now finally return. 



The blocks protruding from the hills ar ranged alon^ the shores of PnU Uhim are more 

 solid and less deconip^ublc masses aud nuclei t of which the formH, and the directions of the 

 sides and a*cs , hate, in uhnont every sn^laiii'ii , Ikch determined by slruclural planes, and 

 which remain after the surrounding rocks hate disintegrated and been washed away. With 

 re-peii Lii 1 1 1 ■ - l.iller , it s-: ubviiui* thai while the Id and has Ii'-i-j* >■•. '■ h i 1 1. In Uip ifrowlh 

 of alluvium in its bays., its mora open coast has been slowly retreating so ibnt what 

 wiis unrc a pari <jf tin l J. m l i> nnw a band an iU border Washed by (he sea but still 

 ejthi biting numerous rocky remnants. The larger masses still evidently occupy their original 

 position-. I i^pieally I heir seaward face \* curved. Somali aies another mass stands behind 

 merely scpn rated frtim that in Front by a rhusm wIida- *td« ore parallrL Willi respect 

 lo the dccnmpoeilion of Ihe rock* on the bills } the soil is entirely derived from this source 

 with ihe exception of a very slight superficial miilure of vegetable matter, which in raany 

 places is absjent. In general r however , the bloeki that remain nrc decomposing with eieectl- 

 ing slowness. One exception 1 noticed in the KW- — SE. side ef the rock described at 



