and the outer bark tends to show a very slight cross- 
fissuring. This is usually a sign among cinchonas of a 
valuable bark, but in this case it is almost completely 
lacking in alkaloids (traces only), recent wartime analyses 
made by the laboratory of the U.S. Government Cin- 
chona Mission in Lima checking with the results ob- 
tained by Howard in the last century. 
In northern Peru, the moist highland forests where 
cinchonas abound (6000-8000 feet), the so-called ‘‘cove 
forests,’’ are isolated from one another by pajonales 
(grasslands or cultivated areas) and are usually limited 
to local guebradas. The ridge-margins of such forests are 
made up of smaller second-growth species which merge 
gradually and deeper in the quebrada with the taller 
dominant species of the forest climax. C. parabolica is 
a tree of the second-growth areas bordering these forests, 
where it is often found intermixed with a weedy variety 
of C. pubescens, called ‘‘amarilla,”’ and at higher eleva- 
tions sometimes with the variety of C. officinalis known 
as /qja. Within the climax forest, often only a quarter- 
mile distant, large trees of colorada, another variety of 
C. pubescens, may grow. Thus C. parabolica is in close 
association, in northern Peru at least, with three other 
cinchonas. Despite this, I have not observed any signs 
of hybridization between C. parabolica and associated 
species of the genus. 
COLLECTIONS EXAMINED: 
Ecuapor: Type of C. parabolica “*‘ex Loxa’’ (localities cited by 
Howard include Laterna, Vileabamba, Valladolid, Sabaneta, Quebrada 
Onda, and Cruz Grande); photograph of type ex herb. Madrid (Field 
Museum negative no. 29638). Type of C. rugosa from the Province of 
Cuenca (Pavon ex Howard). Type of C. Mutisii var. B. from the Pro- 
vince of Loja (Pavon ex Lambert). 
Peru: Amazonas: type of C. Delessertiana from the Province of 
Chachapoyas, Mathews (herb. Delessert, Standley !). A specimen rep- 
resenting C. Delessertiana, in the herbarium of the New York Botan- 
[ 154 ] 
