CHINA, MONGOLIA, AND JAPAN. 93 



Numerous remains of ancient workings, by this method, are found in the neigh- 

 borhood. 



Throughout this region the forest is dense ; among the trees 1 noticed elms and 

 a wild mulberry with black fruit. Fierce, large flies, of two kinds not seen on the 

 sea-shore, swarm in these woods, covering horse and rider, and leaving bleeding 

 wounds wherever they strike. The creek abounds in mountain trout and salmon. 



August l-Jtth. Returning to Kunnui on the sea-shore, we followed the beach to 

 the village of Woshimanbe. 



August 15th. At this village we left the bay to cross over to the west coast. For 

 several miles the road lay over the terrace belt, here covered with drift. At the 

 divide we found a broad, marshy tract through which a large creek winds on its 

 way to the Japan sea. This stream we descended in a small flatboat. 



The prevailing rock across this low part of the ridge was, so far as I could judge, 

 an argillaceous deposit, apparently the same that forms the terraces. 



The forest contained, chiefly, large beech, birch, and maple trees, with oaks and 

 scattered firs, and the usual dense undergrowth of cane. The banks of the streams 

 were lined with water willows. The creeks abound in trout, and the gravelly bot- 

 tom is often nearly hidden by colonies of unio. As we approached the bay of Odaszu 

 the country became more open, and leaving the creek we descended over two ter- 

 races of drift to the village of Odaszu on the sea. 



The southern shore of this small bay is shallow and shelving, with a broad beach ; 

 but the eastern and western sides are rocky, the rocky bluifs descending into the 

 sea, a feature common to all the west coast, so far as we followed it, and indeed to 

 the shores of aU the Japanese islands. 



August 16th. Leaving Odaszu we continued our journey northward along the 

 coast. Here, also, high terraces face the sea, but they are formed of the tufa-con- 

 glomerate formation, the level surface being due to a recent deposit of gravel and 

 sand. This conglomerate is traversed near Odaszu by dykes of a dark gray rock, 

 much weathered, containing crystals of a triclinic felspar, and opalescent chalce- 

 dony. The conglomerate at Isoya is traversed by dykes of an amorphous rock 

 containing crystals of triclinic felspar. 



Near Isoya there is a deposit consisting of beds of sandstone, argillaceous mate- 

 rial, and volcanic ashes,^ with fragments of pumice, and also of the argillite which 

 has been mentioned as occurring at Washinoki and Kunnui with a vermiform 

 fossil. The pieces of pumice contain beautiful double-pyramid crystals of quartz. 

 This deposit is younger than the neighboring tufa-conglomerate, which had suffered 

 much from erosion before the deposition of the beds in question. It continues 

 northward till it abuts against a mass of volcanic rock, that forms the headland 

 south of the mouth of the Shiribetz river. This stream rises nearly north of Cape 

 Edomo, and flows westward through a fine, broad valley. All the gravel brought 

 down by the river seemed to be trachytic detritus. 



' For the interesting results of a microscopic examination of this material, see Mr. Edwards' Letter 

 (spec. No. 11), Appendix 3. 



