20 
the water along* our coasts. Now, in order to obtain a clue to these, 
one looks quite naturally for temperature observations. We 
look in vain, however, for any organised coastal records. It is 
rather strange that the Meteorological Department has not arranged 
long ago for sea temperatures at its coastal stations. We have 
fortunately some data collected by Mr. J. J. East, of Perth. This 
gentleman has worked out carefully the sea temperatures taken by 
the engine-room staffs on various steamers — -mail and coastal — in 
Australian waters. The temperatures apply to the sea water as 
pumped into the condensers, and of course cannot be relied upon as 
quite as accurate as readings made with certificated thermometers 
and the usual oceanographical apparatus. The results, however, go 
some way to till up a big gap and they are quite interesting. The 
coast of West Australia appears to be washed by water the tempera- 
ture of which ranges from about 60°-67°F. on the South to 74°-87°F. 
on the North-West coast. The seasonal change is thus not very 
great. The highest seasonal temperatures recorded in the short 
period of observation occurred in February and March; the lowest 
in August and September. 
At Geraldton, the nearest port to the Abrolhos, the sea tempera- 
tures vary from about 62° F. in winter to 74° F. in summer. 
Now the temperatures taken as one leaves the coast between 
Geraldton and Sharks Bay indicate that the inshore waters are colder 
than those some little distance out, and a map showing the isotherms 
plotted from engine-room temperatures (June, 1911) shows a very 
distinct tongue of warm water passing down the coast, the tip reach- 
ing down to about the Abrolhos. A very distinct zone of cold water 
appears at times lying close to the coast and extending from the 
Leeuwin northwards, even north of Sharks Bay. 
Temperatures taken by me at the Abrolhos also indicate the 
presence of warmer water there than was found in the harbour of 
Geraldton at the same time, although the difference was not half that 
which Saville Kent recorded in 1894. There is, therefore, a con- 
siderable mass of evidence in favour of a tropical current passing 
southward some distance out from the coast and gradually dis- 
appearing as the Abrolhos are approached. This would lead one to 
expect a curious distribution of marine organisms in that area of 
the Indian Ocean, including the Abrolhos Islands, and washing the 
mainland near Geraldton. We hope to be able to map out in some 
detail this distribution. 
Up to date I have little, if any, evidence supporting the German 
contention that the upwelling of colder water along the coast might 
account largely for the marine fauna of the Abrolhos Islands. If 
such a thing took place, caused by the factors suggested by 
Michaelsen, it would in my opinion occur at the Abrolhos Islands 
themselves, for they are situated right on the edge of the continental 
