1898.] Funafuti, Rotuma and Fiji. 483 



and P. brevicornis. These bottles I tixed upside down with open 

 mouths in a pool off Solkopi, where there was a free circulation of 

 water, but no sand or dirt ; at low tide they were covered by 

 4 feet of water and their position was one in which they would be 

 fully exposed to the light. At the end of a week five bottles 

 were happily unbroken, and of these, four had a small amount 

 of gas in each. The gas was then all collected together into a 

 tube, which was next held in the water over a small piece of 

 metallic sodium in a perforated tin box, until twice the quantity 

 of hydrogen gas was added to the other gas. The tube was then 

 removed and, on being turned up, a lighted match was placed at 

 its end and, the thumb being removed, a distinct explosion re- 

 sulted. From the four pieces of Pocillopora, only about 8 cms. 

 of gas were obtained in a week ; the same experiment, however, 

 carried out for a like time with four pieces of Prionastrcva (the 

 species being taken from the rim of the reef, but not distinguished 

 specifically) gave upwards of 15 cubic centimetres of gas, which 

 exploded in a like manner. From Euphyllia gaimardi, over the 

 polyps of which I fixed the bottles without breaking their stems, 

 I obtained an even greater amount of gas, into which I placed a 

 glowing match, which at once burst into a flame. Madrepora, if 

 broken, gave very little gas, but over some of the branches of a 

 species which grows very abundantly on the reef-flat, and in the 

 pools off Solkopi, I fixed bottles ; the gas given off, though not as 

 great as from the Pocillopora, also gave an explosion. I further- 

 more proved that gas was given off from a species of Fungia, 

 and I had traces from several Favia and one Stylophora ; my 

 experiments, however, on these genera were unfortunately 

 shortened by a strong gale in the end of November, 1896, which 

 broke all my bottles. 



Pavonia divaricata and Pavonia cristata, which were also 

 very abundant in the pool by Solkopi and over which I fixed 

 bottles, gave negative results, as also did some experiments on 

 Porites and Astraiopora in the boat channel. Heliopora and 

 Millepora abounded in the boat channel, but no gas from them 

 was obtained, although they were especially carefully managed. 



On the whole the experiments were not altogether satis- 

 factory, as the amount of gas, given off by different pieces of 

 the same species, varied much more than the sizes of the pieces 

 did. All the reef species of Prionastrcva and Pocillopora were, I 

 believe, experimented on and never failed to give a certain 

 amount of the gas, but the different species of Favia were un- 

 certain. It would have been interesting to discover how much 

 of the gas given off was oxygen, by exploding it by means of a 

 battery, but experiments to this end were unsatisfactory owing 

 to the want of proper apparatus. 



