Having dealt with the giant, we now turn to the pygmies; for pygmies are the Patagonian and South Argentine streams in comparison with the Parana of the upper region. The Colorado basin presents a very curious phenomenon, in that it has lost the whole of its upper tributaries. One of these is the aforesaid errant Mendoza, which, with the Salado (the second river of that name) fail to reach the parent stream and end in the Laguna Amarga, a group of salt lakes situated at no great distance from the Rio Colorado. It is certain that, like Central Asia, Patagonia has experienced an immense increase of aridity--probably in comparatively recent times. The Colorado proper is perennial, and when swollen by melting snows from the Andes it is as broad as the Thames at London Bridge. [Illustration: AMONG THE CACTUS.] South of the Colorado we have the Rio Negro. It runs a solitary course through the desert unaided by any tributaries. It is formed by two other streams, the Neuquen and the Rio Limaz, which has its source in the picturesque lake of Nahuelhuapi. Patagonia, it may be added, has numerous lakes, some of great beauty. Other solitary streams, wholly dependent upon the Andes, may be enumerated--the Chubut, the Desire, the Chico, the Santa Cruz, and the Gallegos. On these rivers Burmeister[9] confesses that his information is very imperfect. They are now somewhat better known, but they are still difficult to explore. His remarks upon the Rio de Santa Cruz, concerning which he had gathered more facts than the others, may be given. He says: "Near the ocean it has a breadth of from 5 to 10 English miles, and is bordered by terraces in flights which rise on either side to a height of 500 feet. The surface of these terraces is occupied by broad plains covered with dry pebbles, and among them grow stunted plants and thorny bushes. It is a savage and gloomy land. Further inland, near the source, basaltic rocks appear which approach close to the river, and its bed is strewn with their fragments, which are about the size of a man's head. Huge blocks of granite and palæologic schists are met with only in the neighbourhood of the Cordillera." Darwin made an adventurous voyage of 140 miles up this river in 1834, and describes it with his accustomed acuteness and accuracy.[10] When we come to deal with Patagonia we shall have another opportunity of reverting to its scanty and little-known river system.