SNAPPER BANK. 147 



appeared that it rose from some fire on the main, 

 distant about thirty-five miles, and that its being 

 visible by us was owing to the extr{ me clearness of 

 the atmosphere. 



The observation spot on Rat Island was on the 

 north-east end, which we placed in lat. 3T 4^' 50' S. 

 and long. V 57' 50" West of Swan River. Having 

 completed our work in the harbour, we left, for the 

 purpose of securing the requisite material for the 

 north-east part of this group, which we found to be 

 a detached cluster with deep water between, and to 

 be also similarly separated from the extreme of the 

 group — a small isle about five feet high, composed of 

 sand and dead coral. The average depth surround- 

 ing the islands was 20 and 23 fathoms, being the 

 same level as that of the great flat or plain on which 

 they rest, and which extends out from the main- 

 land, shelving off at the outer edge of the southern 

 part of the Abrolhos almost precipitously to no 

 bottom with 250 fathoms. We now proceeded 

 southward, to examine the opening between Easter 

 and Pelsart Groups, and to complete the ex- 

 tremity of the northern part of the latter. On our 

 way we discovered a coral bank of 7 fathoms, a 

 mile and a half long, seven miles -E. S. E. from 

 the north-east end of Easter Group. We called it 

 Snapper Bank, from the immense quantity of that 

 fish which we found on it. In half an hour we 

 caught more than we could cure, so that it became 

 necessary to stop the sport. This shews what 



L 2 



