25 



a foot to several fathoms in diameter. These balls are sometimes coraposed of curved la- 

 mellar concretions , which always include a harder central mass or nucleus. The spaces 

 between the concretions are filled with granite of a softer nature which decays readily, and 

 thus leaves the harder central masses heaped on each other, or strewed about. Such heaps, 

 or tumuli , hare been erroneously described as rolled masses brought from a distance to 

 their present situation by the agency of currents that formerly swept the surface of the 

 earth. Examples of this kind of structure occur in the island of Arran , Bohemia , the 

 Hartz , the Fitchtelgebirge , and in other countries" (1). 



In the Pulo Ubin rocks the laminar structure is seen well defined and unequivocal. In 

 the globular form it is as regular as that of trappean rocks; and Mr. Scrope's description of 

 some remarkable examples of this structure in a resinous trachyte or pitchstone porphyry in 

 one of the Ponza Islands might be applied verbatim to some of the Pulo Ubin rocks. It 

 is not confined to spherical concretions, however, for, as we have seen, it sometimes occurs 

 in rectilinear zones (2), or on the plane surfaces of cubical masses ($), and, at other places, 

 in irregular variously curved planes (4). In these latter cases it is not improbable that the 

 nuclei are spherical or hemispherical towards their centres, and that the laminae only be- 

 gan to depart from this form, as the expanding nuclei approached each other and prevented 

 further independent development. In such cases it is obvious that the upper portions of the 

 laminae have been decomposed and removed by meteoric or oceanic action, and, sometimes, 

 by both combined. Whether the nucleus in most of these cases where only the upper por- 

 tion is exposed be wholly globular or pass internally into a cylindrical form I am not at pre- 

 sent able to say. 



In a paper of great interest upon the granitic mountain of the Broeken and its »sea of 

 rocks" read before the Berlin- Academy of Sciences on the lSth December 1842, and of which 

 an abstract is given in the first number of the Journal of the Geological Society of London , 

 M. vos Buch refers the external blocks with which the mountain is covered and the con- 

 centric laminar structure of granite bosses in general, to contraction of the mass on cooling. 

 He gives this view a grand application by suggesting that the body of ellipsoidal granite 

 mountains consists, like small bosses, of concentric layers, each repeating the form of the 

 mountain on a diminished scale, — the whole of this structure resulting from the mecha- 

 nical operation of refrigeration. This structure is well marked in the granites of Devon and 

 Cornwall , which have a stratified appearance, the beds conforming to the surfaces of the 

 schistose rocks when these are superineurnbent. Sir H. de ia Beche considers that the laminae 

 or beds probably agree in form with that of the original surfaces of the granite masses after 

 protrusion. He observed at one place alternating beds of a decomposed and hard granite, 

 and he thinks that the difference of original structure may be dué to a tendency of the 



(\) Edijiburgh Encyclopadia , vol. 14 p. 414. 

 (2) Ante p. 8 &c. 

 (3, Antt p. 10 Are. 

 (4) Ante p. 8 , 18 &rc. 

 22 ste DEEL. 1847. D. 



