The rock pierced by the caverns was everywhere of coral formation, and as water freely penetrated through from the soilcap above, the roof and floor were abundantly decorated with stalactites and stalagmites in all their usual fantastic splendour. I noticed that many parts of the floor of the cave were speckled with white spots resembling bird-droppings, on which drops of water were frequently falling from the roof above, and I formed the opinion that the white colour of these spots was due to the drops of water which pattered on them having traversed a portion of the ground above, from which they did not receive a charge of lime salts, and consequently washing clean the portion of the coral floor on which they fell, instead of depositing thereon a calcareous stalagmite. This surmise was strengthened by observing the absence of stalactites depending from the roof in these situations.

Numbers of small swifts, apparently the same species which is common on the island (_Collocalia spodiopygia_), flitted about the vaulted parts of the cave, looking in the torchlight like bats, which at first sight I felt sure they must be, until our native guide succeeded in catching one specimen, which resolved our doubts. We traversed the more open parts of the cave to a distance of about one hundred yards from the entrance; but finding further progress all but impracticable, from the narrowness of the passages, and the quantity of water of uncertain depth to be encountered, we soon gave up the attempt, and were glad to return to the cool and clear atmosphere of the upper air.

Cruise of the 'Alert' — Part 124

During the voyage from Tonga to Fiji, we spent a good deal of time in hunting up the reputed positions of certain doubtful "banks," viz., the "Culebras" and "La Rance" banks, with a view to clearing up the question as to their having any real existence except in the too vivid imaginations of the discoverers. On the 24th of November, when in latitude 24° 25' S., longitude 184° 0' W., we steamed over the position assigned by the chart to the "La Rance" bank, and here our sounding line ran out to three hundred fathoms without touching bottom, thus sufficiently establishing the non-existence of any such "bank." Our position at this time may be roughly stated as some two hundred miles to the southward of Tongatabu. During the greater portion of the day, the sea-surface exhibited large patches of discoloured water, due to the presence of a fluffy substance of a dull brown colour, which in consistency and general arrangement resembled the vegetable scum commonly seen floating on the stagnant water of ditches. This matter floated on the surface in irregularly-shaped streaky patches, and also in finely-divided particles impregnated the sea-water to a depth of several feet. Samples were obtained by "dipping" with a bucket as well as with the tow-net, and when submitted to microscopic examination it proved to be composed of multitudes of minute confervoid algæ. On slightly agitating the water in a glass jar, the fluffy masses broke up into small particles, which, under a magnifying power of sixty diameters, were seen to be composed of spindle-shaped bundles of filaments. Under a power of five hundred diameters, these filaments were further resolved into straight or slightly-curved rods, articulated but not branching, and divided by transverse septa into cylindrical cells, which contained irregularly-shaped masses of granular matter. These rods, which seemed to represent the adult plant, measured 1/2000 inch in width. On careful examination of many specimens, some filaments were observed, portions of which seemed to have undergone a sort of varicose enlargement, having a width two or three times that of the normal filaments. These propagating filaments (if I am right in so calling them) were invested by a delicate tubular membrane, and were filled with a granular semi-transparent matter, in which were imbedded a number of discoid bodies which were being discharged one by one from the ruptured extremity of the tube. These bodies measured 1/1000 of an inch in diameter: when viewed edgewise they presented a lozenge-shaped appearance, and they were devoid of cilia or striæ. A jar full of the sea-water was put by until the following day, when it was found that the confervoid matter had all risen to the surface, forming a thick scum of a dull green colour, while the underlying water was of a pale purple colour, resembling the tint produced by a weak solution of permanganate of potash.

Cruise of the 'Alert' — Part 124