318 PHYTOGKAPHY. 



that may create error or ambiguity, or throw confusion into 

 science. Next in importance is the avoidance of an}* useless 

 introduction of new names. Other considerations, such as 

 absolute grammatical correctness, regularity or euphony of 

 names, a more or less prevailing custom, respect for persons, 

 &c., notwithstanding their undeniable importance, are relatively- 

 accessory.' (Alph. DeCandolle, 1. c.) 



708. The following are universal rules in scientific nomen- 

 clature : 



1. Names must be in Latin or' be Latinized. Those from the 

 Greek (which are more and more abundant, owing to the facility 

 of this language for compounding) take Latin form and termina- 

 tion. 1 Those from modern or other than classical languages 

 should at least have a Latin termination. 2 Hybrid names, 

 namely, those formed by the combination of two languages (at 

 least of Latin and Greek), should not be made. 8 



2. For each plant or group there can be only one valid name, 

 and that always the most ancient, if it is tenable. 



3. Consequently, no new name should be given to an old plant 

 or group, except for necessity. That a name may be bettered 

 is no valid reason for changing it. 



709. Names of Genera are substantive and singular, of one 

 word ; and the same name cannot be used for two genera of 

 plants. 4 They may be derived from any source whatever, from 



1 Thus, words ending with the Greek os generally change it to us, and with 

 on to i/m. A rule not always observed ; for while we have Epidendrum and 

 Oxydendrum, Linnaeus himself variously wrote Liriodendrum and Liriodendron, 

 Rhododendrum and Rhododendron ; and the Greek form now prevails. 



2 In this as in other cases, some exceptions are well established by 

 custom, but they ought not to be extended. The rule as to Latinization is 

 restricted as respects orthography by the necessity of preserving modern 

 commemorative names in a recognizable form. 



8 But we cannot change numerous old names for this fault, such as con- 

 volvuhides, fumarioides, ranunculoides, and scirpoides (though they ought to 

 have been convolculina, ranunculinn,nr\d scirpina) ; and modern botanists have 

 not scrupled to append the expressive and convenient Greek term -oides (sig- 

 nifying likeness) to generic names not of classical origin. Ex. abutiloides, 

 biroides, darallioides, fuchsioides, qentianoides, Jobelioldes, tournefortioides. In Eng- 

 lish, some hybrids will perpetuate themselves, as for instance terminology, 

 centimetre, millimetre, beaitrocracy , &c. 



4 Very many, indeed, are adjectives used as substantives, as Arenaria, 

 Clararia, Saponaria, Impatiens, Trientalis, and even Gloriosa, Mirabilis, &C. 



Some two-worded generic names anterior to Linnaeus, such as Dens Leonis, 

 Vitis Id(ea, Biirsa Pastoris, remain for sections and species, but not for gen- 

 era. When two words are confluent into one, they are not objectionable, 

 as Laurocerasus, Carlemania (commemorating Charles Leman, Carolus Le- 

 manus), &c. 



