560 The Andes and the Amazons, 



tinent in 1867, this vast homogeneous formation along the 

 Great River had not yielded a single fossil. In the words 

 of Professor Agassiz, " Tertiary deposits have never been 

 observed in any part of the Amazonian basin." And it 

 was on this negative evidence mainly that the distinguish- 

 ed naturalist hazarded the conjecture that the formation 

 was drift.* But the banks of the Maranon prove to be 

 highly fossiliferons. At Pebas, near the mouth of the 

 Ambiyacu, as already stated (page 282), I discovered in one 

 of the beds of blue clay, twelve feet below the surface, a 

 multitude of fossil sliells. Below this bed is a seam of 

 lignite, and then another layer of fossils. I engaged Mr. 

 Hauxwell, an English collector, to search for other locali- 

 ties; and in 1870 he reported a large deposit on the south 

 side of the Maranon, below Pebas, at Pichana. The shells 

 were larger and more plentiful than at Pebas, and were 

 found from six to twenty feet beneath the soil. In re- 

 visiting the Amazons, in 1873, I discovered at Iquitos, 

 more than a hundred miles west of Pebas, a still more 

 prolific bed.f Here the shells occur above, below, and in 



* The history of the attempt to find the traces of glaciation in this equato- 

 rial region is short. The Cambridge professor, who had berated other natu- 

 ralists for theorizing without facts, entered the mouth of the Amazons for the 

 first time in his life with the confidence of a prophet, foreordaining bowlders, 

 moraines, strise, and all the other appurtenances of a gigantic glacier. All 

 proved to be imaginary ; j'et the chief and his satellites stoutly kept their 

 original faith. Professor Hartt, after propounding several modifications, the 

 last one being tlie possible glacial origin of the superficial layer, to which the 

 Pebas sliells had driven liiin, finally owns that, " having no evidence whatever 

 of the former existence of glaciers in the Amazons, the question of the gla- 

 cial origin of the valley need not be raised." For evidence against the sup- 

 position of a glacial epoch at the equator, see Annals and Mag. Nat. Hist., 

 1871, vol. viii., p. 297. Keller, in his late exploration of the Madeira, 

 searched diligently for erratic bowlders; but not a trace of the "foundlings" 

 could he discover. "I never believed for a moment" (writes Mr. Darwin), 

 "in Agassiz's idea of the origin of the Amazonian formation." 



t It is very singular that Castelnau and Herndon overlooked the shells at 

 Pebas, since they are plainly exposed ; and still more strange that Mr. Steer, 

 who examined the beds at Pebas and Pichana in 1871, found nothing at 



