APPENDIX. 535 



Mr. Ward had already separated from his fellow-travellers at 

 Barbacena, on his way to the Tocantins, taking the route by Ouro- 

 Preto and Diamantina. And in order to keep together the adven- 

 tures of the little band who left Rio in company, I may give here 

 a short sketch of his journey, before completing the account of 

 the route pursued by Messrs. St. John and Allen. After leaving 

 the valley of the Rio Parahyba and crossing the Mantiqueira the 

 party found itself in the water-basin of the Rio Grande, one of the 

 principal tributaries of the Rio Parana, which, emptying into the 

 Rio La Plata, reaches the ocean below Buenos Ayres. Eastward 

 of this basin, on the ocean-side of the great ridge which bounds the 

 valley of the Rio San Francisco, arise several large rivers, the 

 Rio Doce, the Rio Mucury, and the Rio Jequitinhonha. It was 

 one of my most earnest desires to secure the means of comparing 

 their inhabitants with each other and with those of the great rivers 

 flowing north and east. As will be seen hereafter, Mr. Hartt, 

 with the assistance of Mr. Copeland, had undertaken to explore the 

 lower course of these rivers ; but it was equally important that 

 specimens should be obtained from their head-waters. While, 

 therefore, Mr. St. John and his companion pursued their way 

 across the region drained by the head-waters of the Rio San Fran- 

 cisco, Mr. Ward crossed the mountains, passing from one river- 

 basin into another, in order to examine as many of the tributaries 

 of the Rio Doce and the Rio Jequitinhonha as possible. To him 

 I owe the materials necessary for a general comparison of the river 

 faunae in these different basins. His journey was a laborious and a 

 lonely one. Separating from his companions at Barbacena he kept 

 on by Ouro-Preto and Santa Barbara into the basin of the Rio 

 Doce, which he followed nearly to the point where the Rio Antonio 

 empties into it. This part of the journey gave him an opportunity 

 of making a collection not only in the head-waters of the Rio 

 Doce, but in one of its principal tributaries also. Thence crossing 

 the Serra das Esmeraldas Mr. Ward entered the water-basin of the 

 Rio Jequitinhonha, commonly railed Rio Belmonte on the maps, 



