264 ROBINSON. 



II. A DESCRIPTIVE REVISION OF THE COLOMBIAN 

 EUPATORIUMS. 



In assembling data for a revision of the Eupatoriums of IMexico 

 and Central America the writer was some months ago led to scrutinize 

 the existing evidence as to their possible range-extensions into the 

 northern portions of South America. The questions which arose were 

 briefly, in how far is the Isthmus of Panama a path of plant-migration, 

 or conversely, to what extent has this narrow neck of relatiAely low 

 land, in recent geologic times, or the pre-existing marine channel, at a 

 more remote epoch, acted as an effective barrier to plant-distribution? 



Correlated with this somewhat abstract inquiry-, arises a more 

 concrete one when it is borne in mind that most of the larger South 

 American collections have been studied in Europe, while nearly all 

 recent work upon the Mexican and Central American floras has been 

 done in the United States. Thus it becomes pertinent to inquire 

 whether to any serious extent there has been unconscious duplication 

 in the results — whether, for instance, of the many species and ^'arie- 

 ties described as new from Mexico and Central America, some portion 

 may not precisely duplicate plants present in and perhaps already 

 described from South American collections, which are sparingly if at 

 all represented even in the larger herbaria on this side of the Atlantic. 



At all events, to put classification of the very numerous Eupatoriums 

 north of the Isthmus on a more secure footing, it seemed desirable, if 

 not actually necessary, to give systematic attention at the same time 

 to those of the adjacent portions of South America. For this purpose 

 a provisional revision of the Colombian Eupatoriums was begun in 

 October, 1917. To the usual difficulties incident to tropical work — 

 deplorably inadequate material, scattered types, fragmentary litera- 

 ture, imcontrasted descriptions, faulty and inconsistent records — 

 there has in the present instance been added the impossibility of com- 

 municating with several important European herbaria. However, 

 by piecing together all available bits of fragmentary data, a treatment 

 is here presented, which includes more than three times as many 

 Colombian Eupatoriums as have been heretofore reeonUnl in any 

 single paper. 



It has long been customary in most minor works dealing with tropical 



