2G 



whole to arrange itself in false bed* coinciding with the surface of the erupted mm, kt 

 □ no 1 1 NT place he observed a simiLir atlernatioo of bed*, which , however, in this case ap- 

 peared fo fee at riijlii an^lr* lo (he tearing of the granitic cams in which they occurred. 

 Si-Jifjrl rock ami 'w.un'iu wire seen in another locality an atlemate tones. These phnimnnui 

 appear la be similar to some of ihosc of Pulo Ubin which we arc eoiriderinff. 



Whalercr were the mechanical rami ilium under wlurh the m&m of jpufa L r £i« solidified 

 from n llmd &ta1Cj in Rlttil nduut liial tn no single uniform r- 1 n * rm w t > scribe the phe- 

 nomena which it preson!*. Assuming # as [ think observation rerpiirCJj lhal Mid whole is of 

 ■One contemporaneous origin, we must allow thnt the chemical ingredients of the mass were 

 irrcjjuWSy <1 isl r L h u I <rd . if rot originally, then at a time immediately proceeding sol id i net- 

 Item. Variable mechanical disturbance may have been mainly instrumental in producing locul 

 incqualiiics in chi-mical ami polar action during the transition from □ flu id to a solid jfate J 

 m.'J itiis ini;[ht have Hhs rJFi-i-r or oil ratling certain ekxnetttt la Fjttipjtifaf |.Ln - , .:nl :\ut*- 

 exciting p^icl LLiLikir mill era logical development*. A part ion of the rocks beneath which the 

 granite rote may have been mdtcd inta it ^ and varied iho relative proportions ol tlie old 

 ingredient!; or added new cure, It lh perhaps rash to pursue I his subject without ait exact 

 tlu-mirp] analysis of the rocks under confide ra I ion. But so far us we can take the ascertain- 

 ed general composition of minerals as a ^nide^ we may enquire whether Uic pas^ge of the 

 Same connected maas from a rock in which mica U entirely absent in Is one in which it 

 predominates rices not support the aboTO views? Can tbe frequent suhsfilution of hornblende 

 for mica be otherwise explained? It is true there is a (jrcat similarity id the chemical eom- 

 pf^ilion of (he different minerals which make up fhc recks of the Island, hut some nower- 

 ful force* must have opera led to cause the segregation , in limited Spaces, of certain elements 

 from the surround inn; fluid mass, Mica, is very inconstant in the proportion of ils iugre- 

 dienlip and fbere [ire Species which approach very closely fo hornblende, but eireu ill IheSc 

 the lime of Iho loiter is represented by potash. Where we find the mica ceasing, and horn- 

 blende faking ils place, we must suppnsc lhal in the original fluid or liscid raaa§ lime had 

 been segregated in the space now occupied by Mie latter. 



A tin; urn stance mcnliQiicd by Sir II. dje Li Bacns with respect to the gmnile of Dgtian 

 and Cornwall bears m^Eurjally on this subject, He says that, towards the borders of the gra- 

 nitic tracts t — that is where the granite comes in contact with Uic circumjacent racks,— 

 it allers ils character, passing from a compound of quarts, felspar and mica to a schorla- 

 cccus rock. The mica lir^l disappears,, being replaced by schorl, and then commonly the 

 felspar ako dun |ineari, aud Ihe rock h converted into a corn pound of riuarJ?, and schorl 

 which are generally aggregated in nearly equal proportions. Sir U. n t.a tkcut a.ha men- 

 tion* lhat near the joints by which- the granite is traversed, its character alters, and that it ii 

 fidjoininn; these joints lhat the character of the mass at ils confines become* particularly 

 qua rl rase and scW-laccou*. In the lower spurs of a rjranite chain in Pinatig I remarked 

 tlmt the rock was devoid of that general uniformity throughout considerable (racl.s wlurh 

 distingukhed Ihc central ridpe, It was \nriable in ite composition and Iciture, being; fre- 

 quently coar.NC- grained. Quarts was largely developed, and ir> some pbecs schorl f which I 

 do not recollect baring any where noticed in the ridge, appeared in large crystals and fl^ 



