18 



scientific names of the species encountered since leaving 

 Soputa : 



161, Pako — Sideroxyhn antrridiferum. 



■ 



102, Di,ui<i — Aqlaia sa pindina. 

 l(>r>, Girida — Breyula cernua. 

 170) Ewa — M any if era minor. 



172, Tikina — Schefflera. 



173, ( '!) — Hydnophytum keiense. 



174, Gigino — Adenanthera pavorUna, 



175, Gn^iima — Pavetta platyclada. 

 ITS, lungers — Elaeocarpus sepikanas. 



180, (?) 



181, Simte — Fagraea racemosa. 



Lauranllms sp. 



ls:>, Yandere — Symplocos uggregata. 



Of these, Nos. 161, 162, 170, and 178 are large trees. 

 The fruit of No. 170 is an edible mango. Nos. 16G, 

 172", 175, 181, and 183 are small trees — uudei growth 

 of rain forest. No. 174 is a medium-sized tree with 

 a pretty wood. Nos. 173 and 180 are parasites on 



large trees. The Hydnophytum is a myrmecophilous 

 plant wilh a large tuber which forms the attachment 

 to the limbs or trunk of the host. This 1 liber attains 



sometimes 18 inches across and 8 inches through, it is 



channelled with galleries inside and inhabited by ants. 

 What is the function of the ant in this symbiosis — 

 generally called myi mecophily — is not known. Possibly 

 the function of the tuber is to store nioislnre, and tic 

 ant plays no part in the life of 1 ho plant, except, 

 perhaps, that of an enemy, drinking the stored water. 

 This plant is one of (he arresting objects of the rain 



forest — a <a*otes<|ue excrescence. The other parasite, 

 No. 180, is a mistletoe with a rather showy flower. 



From Wasida 1 went to Ointatandi. After leaving 

 the Wasida <j aniens, the path goes through some third- 

 rate rain forest in a damp flat too wet tor cultivation, 

 and then threads its way through a succession of grass 

 patches broken by narrow belts of rain forest. On 

 reaching the Kuniusi, near where Ointatandi is marked 



on the map — though in reality that village has now 



been moved some 9 miles further down the river — a 



m 



strip of rain forest occurs. This follows the river down 



for 5£ miles until the gardens of another village are 

 encountered, and a re-growth of weed trees has estab- 

 lished i i self. From Ointatandi I went across the 

 Kumusi, following the Ionia path, and looked over tee 

 timber in the hilly country of the Opi River. I found 



nothing of special interest, the growth on the hill slop -^ 

 being poor and the alluvial land scanty and usually 



farmed. The magistrate at torn a advised me not to 



pursue my invest i^ations in the direction of his district, 



he kne of no good forest in that area. I turned 

 back and examined the forest on the Kuniusi already 

 referred to. That this forest remains in spite of the 

 healthy, vigorous population in the vicinity is because 

 the land is low-lyiee; and often inundated. The Kuniusi 

 here is a large river, 300 yards wide, and is 

 subject to verv rapid rises, owing, no doubt, to its 



great length and the fact that sudden heavy rainstorms 



are a feature of the mountain country where it finds 

 its source. These rain storms at the head waters cause 

 rises of several feet at Ointatandi, where there ma\ 



dipterocarp one. Indeed, only two karawas (No. 1S6) 

 were found. The predominant species in number is 



Xo. 5— okamu— with 29.6 per cent., though its averagi 



volume per tree is so low that it only carries a total of 

 12.5 per cent, of the cubic contents. Ilimo, which was 

 unrepresented at Soputa, here makes up 21 per cent, 

 of the total volume. Subtract these 11,656 cubic feet, 

 and our cubic content is reduced to a smaller volume 

 per acre than that found in the forest near Soputa. 

 The Kumusi ilimo nowhere attains the good even, aged, 

 almost pure stands of the Vanapa. Here and there 

 groups of fine trees occur, but, for the most part, this 

 species is well mixed with others. If a name were to 

 be given to this type of wood-land, I think okamu 

 forest would describe it best. As at Soputa, this tree 

 is most conspicuous. The new species number nineteen, 

 as follow : 



No. 184, 



Hangeni. 



— A large tree with a stercu- 



lacious wood. 



Xo. 185, Oeh. 



—A large tree (Elaeocarpus coma- 

 Ins). 



No. 186, Wuwura. — A medium-sized tree with an 



No. 187, 



interlocked timber. 



—A large fig 



Nininge. 



J 



with 



a 



g° 



d 



straight bark. (Ficus sp.) 



No. 188, Nahihi. — A large tree with alternate 



cordate leaves. 

 No. 189, Eehoro. — A large tree with hairy, broad 



leaves. (Ficus sp.) 



No. 190, Gongogo, — A large tree with an umbrella 



crown. 



Scscw a (Barringtonia caly plrocalyx). — 



A very common third-story tree with 



beautiful pink flowers. There 

 is a rarer variety with white flowers. 



—A large tree 



No. 191 



(Polyurias sp.) 



very 



X 



o. 



192 



' 



Mongua. 



(Cri/ptocarya 



sp.), yielding a sound wood. 



A large, 



heavily-buttressed tree. 



No. 193, Busa.— 



N. 194, Au-u-jo. 



No. 195, Onga. — A large tree yielding a soft light 



V large tree. 



(Ficus sp.) 



wood. 





No. 196, Bofere. — A large trae with a brown- 



hearted wood. 



No. 197. 



Ca 



»<tmun 



pineisti [>iujht. 



A 



Xo. 



No. 



198, 



199. 



large tree with leaves. 

 Kiiuina. — A irood hardwood tree. 



. — A large tree with a pink-heart 



wood. It exudes latex. 



No. 200, Ondodo.— A 



large 



tree th 



Q i 

 (1 I 



is hard to 



stand-point, and 



distinguish from 190 from a general 



whose timber has 

 well defined soft tissue like No. 201. 

 Its juvenile leaves 

 length. 

 No. 201, Asawa. — A large tree. The leaves have 



attain 5 feet in 



No. 202 



deep axilla at the junction of petiole 



and branchlet. 



Sirua. 



A 



large 



tree yielding the 



most 



commo 



Northern Division. (Campnosp< > r- 



mum brevipetiolata.) 



narrow- 



have been no rain for several days. The forest is 



from 10 chains to halt* a mile deep from the 

 river bank; its stocking is pom-, and it is only because 



no other area of huge timber was found in this j n 

 that a survey was made of it. The area of forest ot 

 this type is about 2,540 acres, and the area covered by 

 the strip survey was just 100 acres, which gives us a 

 relation of 4 per cent. The data are summarized in by the Motuans for cordage, occurs, but is not used by 



Tables XIII., XIV., and XV. It will be seen that the Binandele people. No. 94 is a hardwood. No. 51 



Of these, none deserves very special remark here. 

 The last, if there were enough of it in any part of this 



country, would be worth attention; it is a good .soft 



wood. 



Of the Galley Reach species not already found in 

 the Northern Division, No. 34 — ilimo — is v 

 important and valuable. No. 74 — vauea — so much used 



the 



most 



ilie species total 47, and are represented by 398, con- (Cinnamomum inassoia) is the cinnamon tree — the same 



species which yielded the product that was exported in 



the old daya under the name of !Massoi bark. Beccari 



taming in all 55,137 cubic feet of sawn timber, or 551 



cubic feet per acre, and I3S . uWc feel per tree. The 



forest, while better than i hat near Soputa, is still tar 



front the Philippine standard, and it is no lon^-r a 



calls this tree M $sj) : <i arowafica. For fuller descriptions 



of the various species, Bee s>etimi VI. The Kumusi Kivcr 



