356 ORIGIN OF ERUPTIVE AND PRIMARY ROCKS. 



in the form of a long drawn mass, in the centre of which it is dis- 

 tinct and characteristic, while towards the hanging wall it gra- 

 duates into gneissoid rock. The central granite or protogineof the 

 Alps, according to Delesse, also graduates at its limits into gneiss, 

 and this, according to Raymond and Charpentier, is the case with 

 the colossal granite masses of the Pyrenees. These various in- 

 stances furnish good grounds for maintaining that gneiss bears 

 the same relation to granite, that diorite slate or hornblende 

 slate bears to many granular diorites, the micaceous selvage of 

 the Kerrera trap vein to its granular centre, and the numerous 

 instances of stratified or banded porphyrites or trachytes, to the 

 corresponding granular rocks. In short there would appear to be 

 reason for assuming that gneiss is as much an igneous rock, as 

 are the banded or stratified varieties of igneous rocks just men- 

 tioned. The instances just given prove at least that certain 

 gneisses are eruptive, because they are nothing else than an out- 

 ward covering, a contact modification of the eruptive granitic 

 masses. There are, moreover, instances on record of gneisses occur- 

 ring in veins, and sometimes enclosing fragments of other rocks. 

 Humboldt mentions an instance occurring near Antimano, in Vene- 

 zuela, where mica slate is intersected by veins from thirty-six to 

 forty-eight feet thick, and consisting of gneiss filled with large 

 crystals of feldspar ; and Fournet maintains that in the mountains 

 of Izeron, true eruptive gneisses occur in veins intersecting other 

 gneissoid rocks.* Darwin relates that the granitoid gneiss of 

 Bahia contains angular fragments of a hornblende rock, and that 

 a similar gneiss occurring in Botofogo Bay, near Rio Janeiro, con- 

 tains an angular piece seven yards long and two broad, of a very 

 micaceous gneiss.f Instances of the same nature have been ob- 

 served by Naumann near Ullensvang in Norway and by Boeth- 

 lingk near Helsingfors in Finland.^ The most satisfactory ex- 

 planation which can be given of the formation of the gneissoid 

 selvage to granitic masses is that which is given by Phillips in 

 the case of syenite, and already quoted. Tt is a consequence of 

 crystallization under restraint or pressure, accompanied by a 

 movement of the solidifying mass somewhat in the same manner 

 as indicated in the case of greenstone. Naumann adopts almost 

 the same explanation in referring to the formation of igneous 



* Naumann, Lehrbuch, II, 180. 



f Geological Observations on South America, p. 141. 



t Lehrbuch, II, 113. 



