observations may be pertinent. The species was first ob- 
served growing just east of Huancabamba at Tabaconas 
(in the Department of Cajamarca) which is some 200 
kilometers (airline) northwest of the Province of Chacha- 
poyas, the type region for C. Delessertiana. Subsequently 
trees were also observed near Bambamarca (Cajamarca). 
Although cinchonas were sought in the Chachapoyas 
region, the species was never seen there. Mathews’ speci- 
men undoubtedly originated not in Chachapoyas itself, 
whose environs have long been depleted of forest, but 
somewhere in the outlying fringes of that province. The 
range of the species extends into the Loja region of 
southern Ecuador, the data of Pavon published by How- 
ard giving as localities ‘‘72 via collis Laterna, in collibus 
Vilcabamba, Valladolid, locis altis frigidis Sabaneta, Que- 
brada Onda et Cruz Grande nominatis, Provincia Laja.”’ 
Thus C. parabolica is known from the Departments of 
Cajamarca and Amazonas (and undoubtedly eastern Pi- 
ura) in northern Peru and extends northwards into the 
Provinces of Loja and Azuay in southern Ecuador. 
C. parabolica is a rather small tree, the largest example 
seen by me reaching a height of about twenty-five feet 
and a diameter of eight inches. The species is perhaps 
the easiest of all cinchonas to recognize, because of the 
very dense pubescence (noted by Standley and others) 
which covers all parts of the young growth of the plant. 
But even more characteristic in fresh material (though 
often lost in pressed specimens) is the strongly bullate 
or rugose nature of the subcoriaceous leaves whose mar- 
gins often inroll or curl under conspicuously. ‘These 
characters possibly account for the Spanish name crespilla 
which is not only used today among cascarilleros of 
northern Peru, but also is cited by Pavon in his descrip- 
tions of C. parabolica and C. rugosa from Ecuador. The 
thin, light-to-dark gray bark of this tree is quite fibrous 
( 1538 ] 
