known to exist. This
channel, which is left white in the chart I am describing, is painted
over in the specimen dated 1550 [see map pp. 68-69], as though it were
blocked, and two men are represented with pick and shovel as in the act
of cutting it open.

Curiously enough, in both maps, the upper silhouette of the landscape in
this part defines the real south shore of Java.

On the continental part, the Australian Alps, the range of hills on the
western and north-western coast, and the great sandy interior of
Australia, are also roughly sketched in. Was it all guess-work?

PLACE-NAMES.

It will not be necessary, I think, to give an elaborate description of
the place-names that occur on this map; those who wish to know more about
them may consult my larger work on "The Discovery of Australia."

We need not dwell either on those that are inscribed along the northern
shores of Java, well-known to the Portuguese twenty years at least before
these maps were made.

The southern shores of Java are joined to Australia, or, at least, only
separated from it by a fictitious river named Rio Grande, the Great
River, which follows the sleek curve of the "pig's back" described by D.
do Couto, the Portuguese historian.

In the Portuguese sphere some of the more salient features of the coast
lines bear the following names:--

_Terre ennegade._ Ennegade has no possible meaning in French.

It is a corruption of Terra Anegada which means submerged land, or land
over which the high tides flow considerably. It refers to a long stretch
of shore at the entrance to King Sounds, where the tides cover immense
tracts of country, and which has, in consequence, been called Shoal Bay.

_Baye Bresille;_ Brazil Bay, corresponds with King Sound.

The islands on the western coast, known as Houtman's Abrolhos,* and those
near Sharks' Bay, are all charted with the reefs that surround them,
although they bear no names on this map.

[* _Abrolhos_ is a Portuguese word applied to reefs; literally, it means
"open your eyes."]

Lower down, there is a strange name, that has led to some stranger
mistakes; it is LAMA, or LAME DE SYLLA, written HAME DE SILLE on another
of these maps. It is a curious jumble that I have not been able to
decipher; it occurs close to the mouth of the Swan River of modern
charts.

Later French and Dutch map-makers took it for the name of an island in
that locality.

Now, in those days, navigators and geographers were c