EASTERN BRAZIL—-CHASE 391 
in the river risky. I saw this fish only on the table, where it is 
excellent eating. 
I made a day’s trip to the Rio Salitre about 45 kilometers to the 
west. The scrub (or caatinga) is much denser and near the river are 
trees, but the earth is badly eroded. I found some interesting 
grasses here, one known as Paspalum denticulatum var. ciliatwm (but 
a good species), which is very rarely represented in herbaria and 
which I found nowhere else, being abundant in low wet ground. 
‘At Joazeiro there is a Horto Florestal, about what we would call 
an agricultural experiment station. It looked very promising with 
fine mango and orange trees, plantations of Eucalyptus, and some 
good Duroc-Jersey and Poland China swine for breeding. 
From the train I saw a stretch of promising sandy savanna at 
Parafuso, a few miles north of Bahia. Returning there I had one 
of the richest days of my entire stay in Brazil. I found several 
little-known grasses I had been hoping for and many more I knew 
in the herbarium but had never seen growing. 
I made a trip across the bay to the little town of Cachoeira at 
the head of navigation of Rio Paragassti. If the Bay of Rio de 
Janeiro were not so famous for its beauty we would hear more of 
Bahia. The bay and its steep surrounding hills, with mangrove 
marshes filling the indentations, is indescribably lovely. 
A giant aroid (Montrichardia arborescens—what Beebe calls 
“muckamucka”) grew in the water at the base of the hills as we 
entered the wide mouth of Rio Paragassi. I hired a man with a 
dug-out canoe to take me out to a dense growth of this in the river 
between Cachoeira and Sao Felix, to get an enormously tall grass 
(Panicum rivulare) growing in it. 
I spent a day in the dry region about Feira Santa Anna and 
another walking back from Serra to Cachoeira, about fifteen kilo- 
meters. A little stream full of rapids and falls kept me company 
much of the way. Here I found Hymenachne condensata, the type 
of which I had seen in Raddi’s herbarium at Pisa and which was 
represented in the United States National Herbarium by a single 
fragmentary specimen. I afterwards found it plentiful in a single 
locality west of Rio de Janeiro. 
I had two more good days in marshes and sandy savannas be- 
tween Alagoinhas and Matto de Sao Joao, about 125 kilometers 
northwest of Bahia, and sailed on January 6 for Rio de Janeiro. 
I reached Rio de Janeiro a second time January 9, midsummer. 
The next day, in company with Dr. Horace Williams, the chief of 
the geological survey, I visited Pao de Assucar. The summit is 
reached by an aérial trolley, the first section of which runs from 
the base to Morro de Urca, the second from the other end of Morro 
