South of the line AB, the rainfall is largely due to a similar low-pressure belt which migrates up from Antarctic regions in our winter, thus following the sun as he moves north to the Tropic of Cancer. Antarctic cyclones and depressions during winter pass regularly along our southern coasts, and give steady rains throughout the colder months.

In the interior the dominant weather factor is the high pressure belt. This always covers some portion of the continent, and here the air is slowly descending and being warmed. Hence it absorbs moisture rather than parts with it, and so these regions are normally dry. Moreover, on the northern side of the high pressure belt blow the permanent south-east trade winds. They blow from land to sea, and from cooler to hotter regions, and both these factors make for drought.

On the east coast, however, the trade winds are on-shore winds, and especially in N. Queensland, they drop great quantities of rain as they become cooled through rising over the coastal ranges. Moreover, a special series of cyclones is found along the east coast, especially in autumn, and these also help to fill in the gap between the summer influences and the winter influences. Hence, from Cairns

(Q.), south to Mt. Gambier (S.A.) (at E in fig. 5), the country is favoured by rain-bringing influences throughout the year, except perhaps in spring. This portion of the continent will always be the most prosperous segment of our Commonwealth.

In the south-west the rainfall, though low, and confined to winter, is very reliable, and exactly suits the requirements of wheat culture.

Perhaps too much stress has been laid on the total amount of the annual rainfall, for the certainty with which the average rainfall may be expected is of equal importance to the settler. For instance, Roeburne (W. A.), Northam (W. A.), Tennants Creek (N. T.), and Cobar (N. S. W.), each have 15 inches per year, but there is little similarity in the value of the falls.

The Roeburne region is marked by the most unreliable rainfall in Australia. In 1891 it received only 0.13 inch, while in 1900 there fell 42 inches. Tennants Creek has a totally dry cool period from April to October. Cobar receives its rainfall uniformly through the year, and there is barely enough in the wheat-growing season for the growth of dry-farming wheat. Northam, near Perth, receives its 15 inches entirely in the winter months, and is consequently much more suited for wheat and close settlement than any of the other three localities.

The heavy line (A to D) in fig. 5 encloses the region where rainfalls are very unreliable. Here any locality will probably vary 30 per cent. from the normal amount, and may, of course, vary to a much greater degree. The average variability rises to 50 per cent. near the west coast and in the centre of the continent. (See Footnote, page 334.)

It has been suggested by some optimists that summer wheats and Indian pulses, etc., would do well in the Northern Territory on the Barkly Tableland, for they say that the average temperature and rainfall are not unfavourable. But the rainfall is not reliable, and this probably explains why this region is less promising than the Victoria River country in the same latitude further west.

Hence, as regards future prospects of close settlement, we can assign approximate values as follows to the regions shown in fig. 5.

Of these regions, 1, 2, and 3 are largely suited for closer settlement; 4, 5, and 6 will probably all be found suitable for grazing. Little can be expected from region 7; its rainfall is too unreliable.

PART VII.

White Settlement in the Tropics.

The ardent White Australian, first of all, usually denies that we have any real tropical region in Australia; secondly, he dwells on the unrivalled possibilities of our northern lands; and, thirdly, he maintains that there is nothing to prevent close settlement in a large part of Australia north of the Tropic of Capricorn.

Unfortunately, comparative climatology gives little support to any of these contentions. I have already shown that there are no highlands in the hotter regions (except the Atherton Plateau), which are worth considering as cool areas for close settlement. If one studies the actual isotherms in the Tropics, it is found that northern Australia contains in the Wyndham region (average, 84.4 degrees) the hottest "wet summer climate in the world. At any rate, I can find no places except around Timbuktu (84 degrees), Massowah (86 degrees), and Tinnivelli (84 degrees) which rival it; and of these only the latter has a rainfall resembling that at Wyndham. Modern data show that the "heat equator passes through Wyndham and Darwin; and, indeed, Northern Australia is one of the hottest tropical regions.

The elusive element of comfort enters very largely into the question of tropical settlement. It is determined physically by the conditions of heat and moisture, and in so far as a single instrument may record comfort, this is best indicated by the wetbulb thermometer. The readings of this instrument depend on the humidity and on the temperature, and a high wet-bulb reading means that the weather is muggy and hot, the most disagreeable of all types to the Anglo-Saxon. On the other hand, a scorching, dry day is usually not nearly so oppressive, while a cool, damp day merely recalls the characteristic weather of our homeland. In neither of these latter conditions is the wet-bulb reading of such a high order.

As to the standards of comfort, as measured by the wet-bulb, there is considerable difference of opinion. There is a general consensus that an average monthly record of 70 or 75 degrees F. wetbulb marks the upward limit of comfort, and I have adopted this in the scale of discomfort given below. (See also Fig. 3.)