6 nUEMA, ITS PEOPLE AND PRODUCTIONS. 



places to the soutli, on tlio western side of tlie Arakan range, the same veins of 

 fibrous quartz and soapstono occur, where no serjientiue is known ; but the 

 rock may exist undetected, as the sjjots are in dense junj;'le, where any close exami- 

 nation of the ground is ditTicult, and the same mineral also occurs near Sandoway. 

 The serpentine in Pegu is a characteristic dark-coloured rock, passing into a 

 gabbro, and cotitaining bronzito as an adventitious niinenJ, and thin veins or 

 partings of golden chrysolite and carbonate of magnesia. The rock is unfor- 

 tunately so much traversed bj' cracks or divisional planes as to be unfitted for 

 ornamental purposes. Serpentine also occurs in the Andaman and Nicobar 

 Islands, of probably the same ago, and in a similar manner as in the Arakan range. 

 TuAonYTK. — The only known locality for trachyte in British Burma is four 

 miles east by north of the village of Byan-gyee, on the Bassein Eiver, foiu'teen 

 miles south of Nga-pu-taw, and thirty miles south of Bassein. Here a mass of 

 trachj'to occurs six feet in diameter, surrounded by Nummiilitic strata. It is known 

 in the neighbourhood by the name ' Kyoiik ta Ion,' and is probably the projecting 

 end of a ' pipe-vein ' of the rock marking tlie point of issue of an old volcanic vent. 

 It is only fifteen miles east of a straight line drawn from Barren Island, an 

 active volcano in the Bay of Bengal, to Puppadoung, an extinct volcano in Upper 

 Burma. 



YOLCANOES AXD MUD VOLCAIS^OES. 



The only volcanic islands, or volcanic cones as they really are, in the vicinity 

 of Burma, are Barren Island and Narkondam Island, the former of which has 

 displayed signs of activity in modern times, whilst Narkondam is densely wooded 

 and appears extinct. These two islands are the termination in the Bay of Bengal 

 of the line of volcanic vents of Java and the Malayan peninsula ; but a further 

 extension of the same line or volcanic belt is indicated by the extinct volcano 

 of Puppadoung in Upper Burma. 



The so called ' mud volcanoes ' of Arakan and Burma belong to a totally 

 different class, having no connexion with volcanic agency junperly so called. 

 The eruptions to which they are subject are due to the escape and ignition of 

 volumes of ' marsh gas ' which issues in a vast number of places in Burma along 

 the Arakan range, together with petroleum and feebly saline springs. Mud and 

 water are freely ejected by these ' volcanoes,' and lumps of stone and shale occa- 

 sionally more or less acted on by the flames, but no scorins, lava, or, other products 

 of volcanic action proper. The principal vents are at Memboo, on the Irrawaddy, 

 and in the islands of Eamri and Cheduba on the Arakan coast, with a few 

 scattered vents on the neighbouring islands. The ignition of the gas is ascribed 

 by Mr. Mallet to frictional electricity, and not to the high temperature of the 

 Bubstanccs ejected, which as a rule are unaltered by heat, jjartiallj' calcined lumps 

 being rare and exceptional. 



In some parts of Prome, where springs issue in an alluvial tract or among 

 argillaceous beds, a pool is formed of deep mud, into which any animal falling 

 is in the utmost danger of being engulphed. These mud holes are especially 

 dangerous in dry weather, when they become caked over, with a treacherous crust, 

 with a soft ]iatch of a few feet in diameter in the centre, but which suddenly 

 yields beneath any heavy animal venturing on, and engulphs it in the semifluid 

 paste beneath. 



The known connexion of petroleum and brine springs in America suggests a 

 few words on their relations in Burma. In Burma the only known localities for 

 petroleum lie within the area of unaltered Nummulilic strata, or of rocks of younger 

 age superimposed thereon. Brine springs, on the contrary, though rising through 

 these younger rocks, are more plentiful in an area comjiosed of altered rock of 

 presumably older character; hence, although in America brine springs may indicate 

 the probability of also meeting with pelrolcuin, yet in Piirma a copious brine spring 



