35 



sparse, is a fact, difficult to ascribe to any other reason 

 but fii.it a larger population existed in the past, and it 

 has migrated to other parts of the Territory for various 

 Causes, possibly the most urgent being that it had ex- 

 hansied the land, created the grass, and could no longer 

 farm it. If any population remains, it is always suffi- 

 cient to maintain the grass lands, m _, 



going hunting will set the whole countryside 

 Outside " ( \vy helts," viz., parts of the country which 

 have two well-marked seasons, a dry and a wet one, 

 all grass lands, except on slopes so precipitous as to 

 make tree growth impossible, are artificially made by 

 human agency. Woodland in the foothills has an even 

 better chance in its battle with grassland than in the 



for a couple of boys 



on fire. 



one of the headwaters of the Kemp Welch, differ from 

 the savannah forests that occur further west, between the 

 JSTaro River, which flows into the Brown, and the village 

 of Kagi. Here Omuwrma nodi flora is a prominent tree, 

 and is to be found in close association with Eucalyptus 

 ft ' rrl Hi ' oni Is; also, curiously enough, it intrudes into the 

 foothill forests, and in one place I saw it thriving 

 alongside intruders from the cloud belt, such as hoop 

 pine, and Pod&e&rpm cupressina, and the two oaks. 



I regret that I can advance no reason for the pre- 

 sence of these areas of natural savannah forests, with 

 their islands of gully forests, at these high elevations 

 in a region which seems truly of a foothill type. 



Geologically, there is, according to Mr. Stanley/' no 

 reason why the foothill timber should not grow on the 

 places occupied by savannah forests, or vice versa. It 

 certainly is not a question of aspect, for they occur in- 

 _ discriminately on all aspects of the foothills; nor is it a 

 way for the better and more permanent foothill forest, case °* dfedmty; the foothill forests occupy as steep 

 In places where the whole population has come down slo P'es as the savannah associations. That the dry belt 

 from its strategical hill top villages to the valleys, the has som ^ influen ce is tolerably certain, for such natural 



low lands, and wherever the grass is left unburnt 

 season a good crop of woody weeds comes up, and these 

 are followed by a whole clan of light-demanding weed 

 trees, which quickly kill out the grass and prepare the 



abandoned sites have entirely reverted to woodland. It 

 is possible that in years to come, as the valleys of the 

 foothills become grass lands, the nath^e will be forced 

 back to the summits of the hills once more, not this 

 time in fear of his neighbours, but in fear of starva- 

 tion. In the meantime his old garden lands are re- 

 clothing themselves with forest. 



savannah types do not occur elsewhere than in the dry 

 belts. So far as the eucalyptus growth is concerned, I 

 have not met it in the N.E. side of the range; but 

 grass areas with cicads in them have been seen at 

 similar altitudes on that side. The eucalypt savannah 

 forests at 4-5,000 feet may be remnants of a much 

 larger area of Australian silva which stretched from 

 the sea coast to the cloud belt, and the foothill may be 



In the dry belt of the Central Division there is a an atrndiag association, ousting the less vigorous form 



Against so pleasing an explanation is the doubt that 



ainereiit type or grass lanas in tne ioorniiis, ana tnese — r>.— — ~w — i™ — *~& < — . i -^ — 



are natural savannah forests, consisting essentially of the eucalypt formation is so degenerate. 



Indeed, of 



grass with a tree growth of Eucalyptus tercticornis, the two forms ^ the foethiK forest appears ' to me Jo be 



with, in certain places a mixture of Casuarina nodi- 



less able to look after itself. The age of the eucalypts, 



to — & ^ 



Where the foothill forests have been de- 



flora and Cicas media. I saw no E. alba, E. jmpuana, 

 or E. clavegera, no Melaleuca or Banhsia; otherwise the 

 general appearance of the savannah forest at 4,000- 

 5,000 feet, along the Mimai, at the foot of Mount 

 Obree, was similar in every respect to the savannah 

 forest above Port Moresby, and on the topi of Ilombron 

 Bluff. I regard these savannah forests as natural, for 

 the eucalypts are of great age — one might call them 

 immemorial. 



st roved for farming purposes, grass land has succeeded 

 it, but no eucalypts have established themselves. Here 

 we see the artificial grass land and the savannah forest 

 side by side, and wherever the native has entirely va- 

 cated the neighbourhood of an artificially created grass 

 patch, the foothill forests re-establish themselves. Up 

 at this elevation there is an undoubted dry belt influ- 

 ence. This is shown, not only by the permanent savan- 

 nah forests of eucalypts, but by the presence of a gully 

 forest, which differs in composition from the foothill 

 forests. These gully forests are a counterpart of the 

 gully forests which occur at low elevations in the dry 



and the wide belt of intervening rain forest and foothill 

 forests, preclure, I think, any possibility of these being 

 a wind or bird introduced species. 



The Mid-mountain Forests. 



This belt of forests is climatically very clearly defined, 

 so far as its lower limits are concerned, for it corre- 

 sponds to the cloud belt. Just so far as the clouds roll 

 down the mountain is to be found the mid-mountain 

 species in forest formation. In actual elevation above 

 the sea, the limit is rather vague, for the cloud belt 

 descends lower in some regions than in others. It is 

 a question of topography and exposure to wind. In 

 some valleys it creeps down to 4,500 feet; while on 

 some slopes it is as high as 6,000 feet. I have laid 

 down the lower limit of this forest region at 5,500 feet, 

 but this figure must be taken as an average one, and 

 not as a hard and fast line. Its upper limit is still 

 less well marked, for its boundary is the mossy forests, 

 where the clouds saturate the atmosphere for the great er 

 part of the time. This varies in altitude for every 

 .lvjnrrn n.f trmnorranViv. so that mv limit of 7,500 feet 



belt, and which I have already described (page, 5). jnust })p 1Ti ,. m i (M i . IS rat l ier arbitrary. Taken by and 

 They differ in composition, but have this similarity: * - —....- 



that' they are permanent islands in the savannah lands. 

 Sometimes a few foothill species crop up, and some- 

 times an Araucaria is to be found rising its high crown 

 far above the gully trees; but, generally speaking, 

 these are exceptions, and the woods that clothe the 

 gullies belong to different species. They are all con- 

 siderably smaller in height and girth than the trees 

 of the foothill forests. When a gully forest is cleared 

 for farming, and later abandoned to nature, the first 



to come up is a dense thicket of Dodonaea 



riscosa, and other light-demanding shrubs, 

 quickly kill out the grass, and prepare the way for 

 the gully species. I did not make a survey of these, 

 as none was of apparent economic timber value. A 

 soft-wooded Albizzia, No. 432, is common and charac- 

 teristic, as is number 374, orena {Rhus umaruhae^a). 

 The savannah forests that occur at the foot ot M0111M 



thing 



large it will not be found far out. It is the factors 

 of humidity, rainfall, temperature, and exposure to 

 or shelter from the wind more than the soil conditions 

 that determines whether mossy forests or mid-mountain 

 forests shall prevail. In the present state of our meteo- 

 rological knowledge it is not possible to set out the 

 limiting factors for each belt. When this can be done, 

 a precise climatic boundary can be set to this and the 



higher floral regions. 



The first thing that strikes the observer rising from 

 the foothills to the main mountain range is that he 

 These has at last left the utterly broken country, and that 



the great mountains are becoming defined. Precipices, 

 deep gorges, subsidiary peaks, do occur, but the main 

 range stands out, and there are long slopes, spurs and 

 ridges leading up to the great divide of Papua. Stand- 

 ing on the topmost ridge of a savannah forest, as one 



Obree, on the sides of the valley of the Mimai, which is 



(a) Reoort OO The investigation of the unexplored mountain regions between 



the Mimai and Kiigi, V,r>:l. Home and Territories Department, Melbourne. 



