130 PKOF. W. J. DAKIN : EXPEDITION TO THE 



both places on the same day, and about the same time. The catches have 

 not yet been examined in detail, but it was fairly obvious, each time, that 

 the plankton of the lagoons contained large quantities of dead organisms and 

 much debris. On the second expedition blasting gelignite was tried as an 

 experiment for the capture of fish. A detonator was pushed into a stick of 

 gelignite, and, after lighting the fuse, the whole thing was thrown overboard. 

 It only worked satisfactorily over the reefs in moderately shallow water, but 

 we could often throw it amidst shoals of coral-haunting fish in these regions. 

 The burning fuse appeared to attract the fish nearer to the explosive rather 

 than frighten them away. After the explosion, numerous fish would appear 

 on the surface. They were not usually dead, but swimming just as if the gas- 

 bladder were distended. Some species did not rise at all, and it would appear 

 that the ascent of the fish immediately after the explosion is not due to death, 

 but to physiological conditions following the shock. 



Another experiment was the use of chloride of lime as a poison in rock- 

 pools. This was suggested to me by Professor Starr Jordan and worked 

 very well, but owing to the tides being rather poor and the wind high 

 on many days when we could have utilized this, it was only tested on a 

 few occasions. 



The Houtman's Abrolhos lie about 40 miles away from the coast of 

 Western Australia, on the very edge of the continental shelf in latitude 

 28° 40' S. (see map, text-fig. 1). The depth of water over the shelf 

 between the Abrolhos and the mainland is very uniform and averages about 

 25 fathoms, whilst the same depth occurs between the different groups 

 of islets. A few miles west of them the sounding-line gives depths of 

 hundreds of fathoms. 



Unfortunately, the Abrolhos Islands are poorly charted and small maps 

 on a reasonable scale deal only with two of them, together with a small area 

 of lagoon in each case which was probably considered to be well sheltered 

 and providing good anchorage. There are no lights of any kind on the 

 islets, and in consequence the fishermen, who have spent years about this 

 area, will not approach close when the sun has set. As a matter of fact, 

 navigation in the lagoons and about the small islets is only safe when the sun 

 is high, and a hand stationed in the rigging can spot the submerged coral- 

 reefs which occur here and there. Coastal steamers give the Abrolhos 

 a wide berth, but more than one ship has finished her life on these barren 

 reefs. In fact, the history of the islands almost commences with a wreck 

 (1629) and the coi-al reef of the Pelsart Island is to-day dotted with the 

 scattered remains of a steamer, the ' Windsor,' which became a total wreck 

 a few years ago. 



Islands in such a position as the Abrolhos — in close proximity to a 

 continent — are full of interest to the biologist, and this is especially the case 

 when they are coral islands. It is not surprising then to find that they 



