50 MISSOURI BOTANICAL GARDEN. 



formed" method of nomenclature. I will enlarge some- 

 what on these principles as I understand them. Assuming 

 undoubted identity and sufficient publication, priority 

 would be decided by the date of the latter, or, in a given 

 work, by lineal sequence, even on the same page. In trans- 

 ferring from one genus to another the plant retains its old 

 specific name under the new genus. But if there be 

 already a plant in this genus bearing the same name, the 

 two cannot be distinguished. Logic would probably require 

 the one having the oldest specific name to stand, but I be- 

 lieve this is not followed by any of the new school. 

 Instead, the one already in the genus is given the 

 advantage, the other taking a new name. Not only 

 should the original specific name be used in all cases 

 but if at any time this has not been done, the original 

 name should be restored. This is the important difference 

 between this system and the first. In the one case, if a 

 new specific name is given, it must be used ; in the other case 

 a new specific name is discarded and the old one restored. 

 It logically follows that the identity of the generic and 

 specific names is no reason for discarding the latter. This 

 gives rise to such binomials as Catalpa Galalpa and An- 

 anas Ananas. It will also give us such names as Coix 

 Lachryma Jobi, L. and Chrysocoma Coma aurea, L., 

 and possibly Melilotus Melilotus officinalis. 



For the same reason, if two species with the same spe- 

 cific name occur in a genus, and subsequently the plant 

 whose name was necessarily changed, is transferred to 

 another genus, the original name should be restored. For 

 example: Celtis Lima, Svv. (Fl. Ind. Occ. i) and C. Lima, 

 Lam. (Diet, iv) are different species. The latter was con- 

 sequently changed to C. Lamarckiana,H. & S. (Syst. vi). 

 But when taken out of the genus Celtis the original name 

 Lima should be restored. 



If two species are united or one is made the variety of 

 another, the older name persists. This is quite inconve- 

 nient when the older is a local or abnormal form. If a 



