1893.] D. Pram — Flora of Narcondam and Barren Island. 59 



■which 2300 fathoms are indicated ; the other in Lat. 12° 30' IS", gives 

 2097 fathoms. These soundings appear to be devoid of authority ; at 

 all events they are quite wrong.* 



A more reliable map is, however, to be found in the same work.f 

 This map, designed by Petermann and drawn by Habenicht, is, unfortu- 

 nately for our purpose, on a smaller scale than Berghaus' map. It shows 

 Carpenter's Ridge jutting southward into the 2000 fathom line imme- 

 diately to the west of the Andamans ; shows comparatively shallow 

 water (between the 100 and the 1000 fathom lines), in the two channels 

 between the Andamans and Sumatra, and indicates a depth of 1137 

 fathoms in Lon. 96° 11' E. and in Lat. 12° 24' N". — practically the situa- 

 tion of Berghaus' 2097 fathom mark ; this sounding indicated by Peter- 

 mann has the advantage of being a real one. Going further into 

 detail however, one finds that not even Petermann's map gives any idea 

 of the true state of affairs within this sea J For the 1000 fathom line 

 is there shewn as enclosing a long and narrow trough half way between 

 the Andamans and Tenasserim ; the three peaks that have just been 

 described are therefore shown as springing from a slope that trends up- 

 wards from the bottom of this trough to the Andaman ridge. Instead 

 however, of indicating a line to the eastward of these peaks the 1000 

 fathom line passes westward between Narcondam and Barren Island to 

 within 30 miles of the east coast of Middle Andaman, where soundings 

 of 1130 and 1159 fathoms have been obtained ; these, it may be remarked, 

 close inshore though they be, have proved (with the exception of a 

 veritable sounding of 1284 fathoms 50 miles east of little Nicobar, and 

 of a doubtful sounding that gives 1260 fathoms with no bottom in Lon. 

 95° 30' E, and Lat. 11° 45' 1ST.) the deepest soundings yet obtained in the 

 Andaman Sea, and are more than 100 fathoms deeper than the deepest 

 indicated along the line that connects Barren Island with Narcondam. 



There is no doubt that taken collectively these three peaks indicate 

 a northward continuation of the line of volcanic activity known as the 

 " Sunda Range," which stretches up from Sambawa and Flores through 

 Java and Sumatra at least to Barren Island. Von Buch in his work on 



* In a previous paper (Joum. As. Soc. Beng. lx. pt. 2, p. 284) the writer was 

 misled by these soundings, which he supposed to have some foundation, into giving 

 the depth of the Andaman Sea as over 2,000 fathoms. 



t Stieler's Hand Atlas, Sheet 58, dated 1884; scale 1 : 30,000,000. 



X In criticising these maps the writer would wish it understood that it is from 

 no desire to cavil that he points out their defects ; it is only because they are worthy 

 of criticism that reference is made to them. Except the Admiralty maps, which 

 are above reproach, no English map with which the writer is acquainted deserves 

 to be mentioned alongside of those in Stieler's work. 



