part 1] THROUGH THE ANDES OF PERU AND BOLIVIA.. 5 



changes, however, have probably been underestimated, and I can 

 find no conclusive evidence to show that the elevation has been any 

 more rapid here than on the coast. 



The occurrence of marine shells at a considerable height above 

 the sea cannot always be taken as conclusive proof of recent uplift, 

 for they are often carried far inland by sea-fowl, and their remains 

 are sometimes so numerous that they appeal- to be in situ. Another 

 case, one in which human agency is involved, is the accumulation 

 on the high-level pampa of shells thrown from the windows of a 

 passing train ; one species in particular, a delicacy much esteemed 

 by the native population, is consumed daily in large quantities on 

 the journey to Arecjuipa. The shells become incorporated in the 

 drifting sand, and are often carried some considerable distance from 

 the line, when their discovery is at first very misleading. 



FromTambo the line rises in a series of curves somewhat steeply 

 to Cachendo (kil. 55, alt. 8250 feet). Beyond this point the 

 aspect of the country undergoes a remarkable change, for the rocky 

 hills bounding the coastal region become buried beneath a vast 

 desert known as the Pampa de lslay,which, although broken through 

 by isolated rounded hills of granite in the neighbourhood of Huagri 

 (kil. 70), stretches as far as Vitor (kil. 122, alt. 5350 feet). 

 Scattered widely over its surface are magnificent examples of 

 crescentic sand-dunes, the white sand of which contrasts strongly 

 with the general reddish colour of the pampa. These are 

 particularly well developed in the neighbourhood of La Joya, 

 where they are frequently 20 feet or more in height. The 

 prevailing winds, during the first half of the year at least, come 

 from the south-east, and the horns of the crescents point in conse- 

 quence uniformly towards the north-west, in which direction the 

 dunes are said to advance at the rate of 100 feet in the course of 

 a year. 



After leaving Vitor a steep winding ascent is made, through the 

 hills known as the Cerros de la Caldera, to the town of Arequipa 

 (kil. 172, alt. 7550 feet), picturesquely situated in the centre of a 

 well-irrigated and fertile plain. 



Throughout this ascent the railway is bounded on its northern 

 side by the deep gorge of the Uclmmayo River, which may be 

 regarded as forming the dividing-line between the foot-hills and the 

 main volcanic peaks of the Western Cordillera. 



The panorama of these giant snow-capped mountains, apparently 

 towering above the city, although in reality some 10 miles distant, 

 is one of extraordinaiy beauty. On the north-west lies the jagged 

 ridge of Mount Chachani (19,300 feet), on the south-east the flat- 

 topped Pichu Pichu, while in the centre rises the symmetrical cone 

 of El Misti (18,619 feet). 



From Arequipa the railway takes a bold sweep to the north-west, 

 so as to circumvent the foot of Mount Chachani. We chose the 

 more direct route, however, between this mountain and El Misti, 

 following the old Inca trail up the right or northern bank of 

 the River Chili, which here cuts a deep gorge between the two 

 mountains. 



