36 Prof. Milne — Across Europe and Asia. 



southern parts of the volcanic districts the rocks appeared to 

 be more felsitic in their character. Some of these were markedly ■ 

 porphyritic, while at Koorstelroi they were phonolitic. The rocks 

 which form the lower edge of the plateau near Kalgan are described 

 by Pumpelly as being porphyritic felsites. Those which I saw in 

 the upper portions of the pass at this point differed from those lower 

 down; while the upper ones were amygdaloidal and somewhat 

 basaltic in appearance, the lower ones were felsitic. "Where these 

 volcanic rocks occurred, they generally stood up above the surround- 

 ing country as hard rough ridges. When they were elevated and 

 brought to view by refractive agencies producing the Fata Morgana, 

 which was a common phenomenon while travelling in the desert, 

 they appeared as rugged islands floating on a level silvery sea. 

 The other rocks which I saw were chiefly granitic. The first of 

 these occurred as boulders which I half suspected had been 

 transported by the action of ice. This I afterwards found not to be 

 the case, as I shortly afterwards (Dec. 28th) came upon a district 

 where, by the ordinary processes of degradation, these boulders were 

 being weathered out from their parent rock. In this district, througli 

 the disintegration of the granite, the road was very gritty and the 

 ground hard. Apparently as a consequence of this, there was an 

 absence of the small burrowing pika, which before had been visible 

 at every point along the road. The buzzard, which feeds upon the 

 pika, was naturally absent also. Not only would the geological 

 nature of the district in this way affect the animals, but it would 

 also have an important influence upon the plants. 



In one place, on the 31st December, I noticed a syenitic rock. 

 Besides these volcanic and granitic rocks, there was the covering of 

 sandy drift. In the Kalgan Pass this had the appearance of a reddish 

 and sometimes whitish marl. As to whether the marls and sands 

 of the Gobi plateau have any connexion with similar deposits in 

 Siberia, I should be afraid to say before further evidence can be 

 adduced, all that I can do being to refer to what I said about the 

 high levels of the terrace deposits in the valley of the Tunka. 



Mr. Pumpelly, who traversed this same region in 1864, at a similar 

 season of the year as that at which I crossed it, was more fortunate 

 in his observations, this being partly due to his having taken a some- 

 what different route. 



Starting from Kalgan at Lake Bilika Noor, he found horizontally 

 stratified — 1. Yellowish grey limestone ; 2. Thin beds of clay or 

 earth with manganese concretions ; 3. Bed of crystalline gypsum ; 

 4. Gypsum with red clay. The last mentioned being the oldest or 

 lowest in the series. This was before reaching the Mingan Hills, 

 where quartzite, compact sandstone, and a talco-argillaceous schist, 

 dipping at a high angle towards the N.E., were found. Beyond this, 

 beds of limestone were met dipping S.S.W. Still further on the 

 road, a white calcareous sandstone, with thin beds of an inter stratified 

 arenaceous limestone, was observed. Sandstone was next found, and 

 then granite, until the valley of the Ulanoor was reached, where 

 clayslate was met with, and afterwards more granite. 



