AMERICAN NOMENCLATURE. 215 



of nearly all post-Linneaii specific names. For very few of them 

 can be so securely established that they will be free from constant 

 danger of being displaced by the discovery, in some obscure work, 

 of slightly older names used perhaps under remote genera. Added 

 to this inherent lack of stability, the unqualified adoption of the 

 first specific name leads to indefiniteness through the constant 

 endeavour to base our nomenclature upon more and more remote, 

 fragmentary, and obscure descriptions of the past, such as those of 

 Rafinesque, while in general the first correct combinations, having 

 been formed in more recent times, when generic and specific limits 

 were better understood, have been based upon or accompanied by 

 fuller descriptions, forming a much sounder foundation for nomen- 

 clature. For these reasons it seems best to adopt the principle of 

 priority under the genus, the whole question of determining in 

 individual cases the proper scientific name being thus greatly 

 simplified, since all competing names are under the same generic 

 designation. It is to be emphasized, however, that this ruling does 

 not lessen the obligation of botanists of the present and future, in 

 making a transfer of a species from one genus to another, to pre- 

 serve scrupulously the specific name without alteration, except in 

 the case of an existing homonym. 



4. The varietal name is to be regarded as inferior in rank to 

 the specific. The variety is the least definite category of classifi- 

 cation, and varietal names have not only been treated witii much 

 greater laxity than the specific, but are generally unindexed, so 

 that it would be a work of years to collate them. To bring them 

 (as advocated by certain recent works) into active competition with 

 specific names would thus tend immeasurably to increase the diffi- 

 culties of an ultimate settlement of specific nomenclature. The 

 rule that a variety may not hold the same name as a species in the 

 same genus is highly arbitrary, and would lead not only to the 

 renaming of thousands of varieties, but the practical impossibility 

 of u^sing in large genera like Aster, Solida[/G, Senecio, Solanum, and 

 Carex any telling descriptive names for varieties, since all such 

 have long since been used for species. 



No specific name should be altered because of pre-existing 

 varietal names for the same plant. Nevertheless, it is recommended 

 as a working rule that whenever a variety is raised to specific or a 

 species depressed to varietal rank the name should be preserved 

 whenever possible. 



6. The principle of "once a synonym always a synonym," 

 while recommended as an excellent working rule for present and 

 future, may not justly be made retroactive. 



Mr. L. H. Bailey's signature is qualified by the following 

 reservation: — "As a statement of the principles or theory of 

 binomial nomenclature, I concur with the above argument, but 

 I am unwilling to subscribe to any code until it shall have been 

 carefully considered by representative assemblies of botanists of 

 the country or the world. Binomial nomenclature is but a special 

 form of language, and all permanent progress in language, as in all 

 other human institutions, is known to be the result of an evolution 



