THE KI ISLANDS. 
109 
—one of those remote and quiet places which are becoming every day more rare, where man 
contrives to live, and to live happily, without railways, telegraphs, and daily newspapers. 
We left Ki Doulan on the morning of the 26th September, and our track for the next 
two days lay amongst that chain of little-known and rarely-visited islands which connect 
Ceram with Timor Laut, and form the eastern boundary of the Sea of Banda. Such are 
the Tionfolokker and Nusa Tello Islands. Some of these are hilly, without, however, 
attaining any great height; some barely rise above the level of the sea; but all are covered 
with a dense vegetation. About sunset on the 27th we passed the southern end of Kanalur 
Island, near enough to observe lights or fires on its western shore. These islands, 
I believe, have never been explored by the naturalist. They could easily be reached from 
Banda or Amboyna with the assistance of the Dutch authorities—always liberally disposed 
towards the scientific traveller. 
At daylight on the 28th we sighted Bird Island, the actual position of which we 
found to be about twenty miles further east than laid down in the charts. The day’s 
sounding indicated a depth of 2800 fathoms, and the temperature of the water was 
ascertained to remain stationary at 37°.5 F., or 3 0 C., from 900 fathoms down to the 
bottom. This seems to prove that the Sea of Banda forms a separate hollow or basin of 
great depth, cut off by submarine elevations from the colder bottom-strata of the Indian 
Ocean in the south and of the Pacific Ocean in the north. Shortly after 2 p.m. we observed, 
for the first time during the voyage, the phenomenon of a waterspout. The day was very 
fine; there was an almost perfect calm, a few showers passing in the distance. Neither the 
barometer, which stood at 30.0 in., nor the thermometer, which indicated a temperature of 
29 0 C. of both air and surface-water, gave any warning of the existence of an atmospheric 
disturbance. The waterspout was visible at a distance of several miles to the northward, 
and, as all the representations I had hitherto seen of similar phenomena seemed to have 
been drawn from memory and very much exaggerated, I took advantage of this opportunity 
to make, with the help of the telescope, an exact sketch 
of what we saw. From the horizontal base of a large 
cumulus cloud, a column or massive thread was seen to 
detach itself, and remain suspended over the surface of 
the sea, sometimes hanging down nearly straight, but 
generally assuming the undulating form of the letter S- 
Being in the shadow of the cloud from which it issued, 
the column was of the same dark neutral tint as the 
base of the cloud, and it stood out with a well-defined outline from the sunlit hoiizon behind 
it. After emerging from the cloud in the shape of an inverted cone, or like the open end 
of a trumpet, its diameter gradually decreased, and for a short time its lower extremity 
seemed to connect itself with a cone of water or spray which rose from the surface of the 
sea to meet it. The phenomenon lasted about a quarter of an hour, and ended by a gradual 
retreat of the lower extremity towards the cloud, until the whole column was once more 
