30 ROBINSON 
state) a subcampanulate rather than cylindrical form. This suggests 
a transition toward the Critonia-group of Sect. Subimbricata in which 
also the leaves are commonly marked with round and oblong pellucid 
dots. 
Il. THE EUPATORIUMS OF BOLIVIA. 
Tue botany of Bolivia is an essentially recent subject, largely 
created by the activities of the present generation. The plants of 
the other Andean countries, Chili, Peru, Ecuador, and Colombia, 
were in some degree scientifically known even in the eighteenth 
century. A good many of them were described in the works of 
Linnaeus and Lamarck. Indeed, before 1800 Ruiz and Pavon were 
already publishing their elaborately illustrated folios on the flora of 
Peru and Chili. In regard to the plants of Bolivia, however,—or as 
it was formerly called, Upper Peru—the records prior to 1890 were 
exceedingly scanty and inaccurate. 
Of the genus Eupatorium, so abundantly represented throughout 
tropical America, DeCandolle in his treatment published in 1836 
(Prod. v. 141-186) included but a single species from Bolivia, namely 
his own E. Pentlandianum, and this appears to be the sole mention of 
the Bolivian occurrence of the genus prior to 1857 when Weddell, 
Chlor. And. i. 216-218, among his Andean Compositae, described as 
new three species of Ewpatorium from the western part of Bolivia. 
In 1865 there was published (Bull. Soc. Bot. Fr. xii. 79-83) a con- 
siderable list of plants collected by Gilbert Mandon, chiefly in the 
neighborhood of Sorata. The Compositae of this catalogue had been 
identified by Schultz-Bipontinus and among them he enumerated 
twenty-two species and varieties of Eupatorium including no less 
than: thirteen which were named as new to science. Unhappily not 
one of these novelties was provided with a diagnosis. It is true, 
some happened to be mixed with other material, distributed under 
identical numbers in Mandon’s exsiccatae, and in order to explain 
their identity Schultz in a few instances mentioned one or at most 
two salient features, such as the number of the florets or length of the 
petiole; but in no case was the description sufficient to establish 
the validity of the species. The Compositae of this list were some 
months later re-enumerated by Schultz (Linnaea, xxxiv, 527-536) 
with a very few supplementary notes. This catalogue of the Mandon 
numbers, having given the earliest clue to the identity of many 
Bolivian plants is of considerable historical importance 0 relation 
