DESCRIPTION OF THE EXPEDITION. idl 
of 2170 fathoms, 44 miles north of Providence reef. As the machine on the port side of 
the Sealark had broken down earlier in the cruise, this accident left us dependent on 
a small hand-machine situated aft, and we were obliged to abandon all further deep 
sounding for ten days, being meantime severely handicapped in dredging on account of 
the slow rate of sounding. At the time it threw some hours on our hands, since we did 
not wish to arrive in the night, and so enabled us to take a few hauls of the pelagic 
fauna in spite of the heavy sea running. We arrived at Alphonse on Oct. 7, and as we 
expected to spend Sunday, Oct. 8, there, did not go ashore. Mr. Alexander and the 
doctor, however, went off in a steamboat, bringing us back specimens of its rocks, It 
was fortunate they did so, as our first anchorage to the south-east proved so dangerous 
_ that we had to shift, losing an anchor in the process owing to the rough nature of the 
bottom. Our second position in the passage to the south between Alphonse and Bijoutier 
was no happier, as, when the tide began to turn between 5 and 6 o'clock, we experienced 
a rapid change of current. It was preceded by a series of short, steep waves, twice 
repeated, which, had it not been for the coxswain, would have swamped the steamboat, at 
that time made fast to the boom. We let out more cable, but soon began to drag ; 
finally, as darkness came on, we drifted into deep water to the east, where we lay-to, 
steaming dead slow head to wind, as we wanted to swing for variation in the morning. 
It was a disappointment, but the rocks secured left no doubt but that Alphonse was 
simply a sandbank, probably of wave and wind formation. Bijoutier and Frangois were 
similar, and there were other banks awash on their reef. Within the reef to the south 
of Alphonse Mr. Alexander found a large lagoon of 6 to 7 fathoms in depth, and the 
half-caste manager of the island, who came off to us in a pirogue, confirmed the existence 
of a similar lagoon between Bijoutier and Francois. The islands belong to the same firm 
as Providence. 
At dawn on Oct. 9 we found ourselves to the south of Poivre in the Amirante Group, 
having decided to avoid Marie Louise and de Neuf, its southern islands, owing to their 
lack of suitable anchorage, the only one offering any protection having been ruined by a 
guano-steamer, which took fire and foundered. That day was devoted to dredging on 
the bank, mostly in about 30 fathoms, the ship finally coming to anchor in the 
evening off Poivre Island (fig. 41; see Pl. 18). 
In the morning we went ashore with the doctor and paymaster, passing over a shallow 
stretch to the north-west of the island, mainly sand with patches of weed and a few coral 
colonies. To the north of the northern island we found a rather tortuous passage to the 
‘settlement overa broad reef. It was low tide and we wandered over the flat, examining 
a series of rocks, which extended along in lines more or less parallel to the present shore. 
They were all of sand and coral formation, reddened by guano and partially recrystallised. 
No doubt they were once part of the island, which must in this position have lost a 
breadth of upwards of 100 yards. At the base of one rock we were interested to observe 
a large stickleback (Lstwlavia), which was darting in and out of the rock, using its pipe- 
like mouth for seizing the worms as they pushed out their heads. The settlement behind 
was of the poorest description, under a half-caste, though the islands were as rich in 
coconuts as any we saw on the cruise. The whole was overshadowed by a high clump of 
SECOND SERIES.—ZOOLOGY, VOL. XII. 22 
