157 Umi River Another, worse, failure of personnel took place this afternoon. The sleeping place of about a dozen of the very rare bat Otomops had been found in a hollow branch high on a big tree in the rain forest. All hands, including the cook and one of my boys, went out after lunch to cut the tree down and collect the bats. The cutting took about an hour. Van posted himself and some boys near where the head of the tree was expected to fall. John took up a position from which he could see the hole in the branch and shoot escaping bats. What happened after that took a little finding out. While the cutting was going on several of the bats crawled out and moved about on the branch. John could have shot them but for some reason did not. Finally, as the tree began to fall, all the bats took wing. John tried to shoot then, and found he had not loaded his gun. Van has Harris' tree-climbing kangaroo, D. matschei, bought for 2 pounds, the price Harris paid a native for it. Saturday, November 21. Day almost completely overcast. The usual onset of light rain about 5 p.m. Rain ended early -- about 7 o'clock. Spent the day on specimens in camp. Boys collected a few things growing near by, then hunted insects. Good catch of Odonata, at the cost of one broken killing bottle and one broken net. Nothing in over 130 fresh trap sets. One Nyctimene in the several mist nets. A Dobsonia shot in a big fig tree on the road near Waterbung. I found this feeding tree by chewed-up remains of the fruits littering the ground. John walked to the village about an hour above the road on the north branch of the Umi to talk Tachyglossus to the natives. They know the beast and say they will go after it. This will be our last predictable chance to collect it, for the talk is that it occurs only in the mountains. So far as we know, only a mandible and bit of skin have been collected in NE New Guinea. Sunday, November 22. Only a little sun today and a change in the weather pattern brought a shower about mid afternoon; no other rain. The weather still comes from the SE. Cloudiness seems mainly on the mountains on both sides of the valley. The valley itself is relatively clear. Botanized in a relic forest patch on the edge of the grass plain about 3 miles back along the road towards Water Rice. Area of 3-4 acres of a drier type of forest than on the hills and in the gorge near camp. Prominent trees Pterocarpus indicus, Sterculia sp., a big Dysoxylum also in the river and hill forests, and in the lower layers abundant Arenga and a small cream-flowered Syzygium. Details in small notebook. Saw indications of both gain and loss of ground by this small patch of forest. All the grass around it had been burned this year. On the SE side the fire had eaten into the forest 10-12 feet on a face and two fair sized trees had been killed. On the sheltered NW side the fire had burned into the brushy edge, exposing a cycad, a true grassland plant, which had survived when the brush invaded the grass. Here and on the savannas of Water Rice a big staghorn fern, not seen in the main forests, is common. Looks like Platycerium grande. Only one Melomys in traps. Natives from Tsingitrompon (r rolled), the village on the Ofim River (N branch of the Umi) visited yesterday by John, came to the rescue with 33 Rhinolophus cf. cervinus of various ages and coloration