a Olacial Epoch at the Equator. 301 



siderable size, particularly tenuis and erectus ; a specimen of 

 the latter before us measures 2 by 2^ inches, and is packed 

 with clay crowded with P. ohliquus. All the specimens are 

 remarkably perfect, except Bulimus and the unknown bi- 

 valve. The valves of the Pachydons are seldom separated, 

 and scarcely ever broken, and none of the shells show the least 

 abrasion. The Neritinaj P. tenuis^ and P. carinatus retain 

 the epidermis, the first displaying various patterns of coloured 

 zigzag lines. Many species, as Iscea lintea^ Liris laqueata^ 

 and Dyris gracilis^ are exceedingly delicate, yet perfect. But 

 Agassiz says the Andean glacier must have ploughed the 

 valley-bottom over and over again, grinding all the materials 

 beneath it into a fine powder. How did these shells escape 

 during " the kneading-process the drift has undergone beneath 

 the gigantic ice-plough?" The supposition that they may 

 have been washed in from another locality must be rejected ; 

 for they are plainly in place, and none are water-worn. " It 

 seems clear," says Conrad, " that they were not transported 

 from a distance, but lived and died in the vicinity of the spot 

 in which they are found." The shells are filled with the 

 same bluish or drab sandy clay, " holding minute scales of 

 mica, and frequently ferruginous," in which they occur. The 

 Pachydons abound in the indurated and concretionary as well 

 as soft parts of the formation. 



Here, then, we have a large collection of shells from locali- 

 ties thirty miles apart, exhibiting seventeen species, all ex- 

 tinct, belonging to nine genera, only three of which have 

 living representatives. The beds, therefore, cannot be later 

 than the Pliocene. There is not one strictly marine genus ; 

 Gabb's. Tellina turns out to be the young of P. tenuis. The 

 deposit was probably of brackish-water origin. Only one 

 specimen of the land-shell Bulimus was found ; and this was 

 about the Only one in the collection which appears to have 

 suffered fracture before deposition. The fact that all the parts 

 are so orderly laid down (lignite, clays, and sandstones) 

 points to a quiet formation, and not to a tumultuous flood or 

 debacle. Any subsequent oscillation must have been conti- 

 nental ; for the beds are without sign of being unequally tilted 

 or dislocated. 



It is quite plain that the drift theory of this formation must 

 be abandoned ; but Professor Hartt, to whom science is in- 

 debted for many minute and careful observations on the 

 eastern border of Brazil, has propounded a new version. He 

 thinks that the clays and sandstone are very late Tertiary 

 and marine, while the superficial unstratified deposit, covering 



