PLANTS COLLECTED IN BAHAMAS, ETC. 55 



Authors are not infrequently cited for varietal names for 

 which they are not responsible; for example, in DC. Prod. 

 iv, p. 541, we find " Spermacoce verticillata , var. Ameri- 

 cana, L. Spec." and " var. Africana, L. Spec' Linnaeus 

 merely says under S. verticillata, " Habitat Jamaica, 

 Africa." (I will remark here that it is much easier to find 

 errors in another's work than it is to exclude them from 

 one's own.) 



Another doubtful point is illustrated by Echinopepon 

 cirrhopedunculatus, Rose, Contr. Nat. Herb. Vol. i, No. 4. 

 In a foot-note the remark is made that some authors would 

 name the plant Echinocystis cirrhopediinculala, others 

 Micrampeles cirrhopednnculata . In case these latter com- 

 binations are used should Rose be cited as their author? 

 Probably not, as the author of a combination thus records 

 his opinion as to the name the plant should bear. 



It would be convenient if, as some have suggested, the 

 names of well marked varieties were always different from 

 all species of the genus. In a critical work their proper 

 relation to the species would be shown, but in ordinary 

 references they could be counted as species. This saves 

 time and is just as definite. 



In citing authors for species which have been separated 

 from others, it would certainly be more indicative to con- 

 nect with both the name of the author who made the separa- 

 tion. A reference to Epilobium color a turn, Muhl. may 

 mean E. adenocaulon, Hauskn., but a reference to E. 

 coloratum, Hauskn. would leave no doubt. This, how- 

 ever, is not in accordance with the new system. 



In case the spelling of a generic or specific name has been 

 changed, should the author of the correction be cited? 

 He certainly is not the author of the genus or species, only 

 the author of the correction, yet it would hardly do to cite 

 an author for a spelling which he did not use. In this 

 catalogue I have used the original spelling so far as that 

 could be ascertained, except in cases of clearly typograph- 

 ical errors. 



