474 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 



taut region, tliis also with its various forms diverging, it may be, 

 along quite different lines; and to this fast-growing "polymorphic 

 species " may have been added a third and a fourdi form, each with 

 its attendant variations. If the herbarium material at hand is con- 

 siderable and represents forms from a wide geograpliic range, the 

 chance of determining with certainty the limits of the several species 

 previously merged or confused is, naturally, impi-oved ; yet also the 

 greater is the need of study and of a thorough search of literature in 

 applying the names correctly. With only scant material available 

 a brief diagnosis of a Jama lea u or a Martinique plant might be 

 tliought to apply well enough to Mexican specimens, in the al)sence 

 of specimens from the type region, and oue name might be made to 

 api)ly to the wdiole. If this happened a century ago the reference, 

 whether right or not, may have the weight of monographic "au- 

 thority " of the intervening period, and the present-day writer may 

 be confronted with the need of determining the boundaries and re- 

 lationship not of one, but of two, three, or more species, as the case 

 may be. Oi', the original form having been rare or not much collected 

 since, the name given to this may have become fixed definitely upon 

 a single species, but this very different from the original and pos- 

 sibly from a distant region. On the other hand, the older Avriter, in 

 the lack of connecting forms since collected in abundance, may have 

 recognized far too^ many " species." When, finally, the limits of the 

 several allied species have been made out, it may appear that in the 

 case of any or all of these which may have received names, the nomen- 

 clatorial type, having been determined not by selection but (often of 

 necessity) by mei-e accident of first discovery, is not truly typical of 

 the species and represents one of the outlying forms. Under these cir- 

 cumstances the nonavailability of a type specimen for comparison or, 

 at least, a lack of knowledge as to its exact origin, becomes a doubly 

 serious handicap. 



It is not to be supposed that the difficulties mentioned e:sist in a 

 study of the ferns alone, though it must be admitted that this group 

 in particular has suffered radically diverse treatment at the hands 

 of various students. This has resulted naturally from the circum- 

 stances. Fern species, partly by reason of their ready dispersal and 

 their imusual breadth of variation, are commonly supposed to occupy 

 extensive ranges that would at once be discredited for the great 

 majority of phanerogams. Undoubtedly a very wide distribution 

 is to be ascribed to many tropical ferns, mainly lowland species, and 

 very many others in one form or another too close to be separated 

 are known to stretch over half a continent; yet marked exceptions 

 occur, as in the Cyatheaceae, which have been found to be relatively 

 local in range and in which, as might be supposed, definite though 



