Garman.] 168 [Marchl9, 



eating the sources whence they are derived. Unless three author- 

 ities are given with trinomials, four with quadrinomials, and so on, 

 the work of tracing the history is greatly increased and becomes 

 practically impossible to anyone except the specialist. Such a 

 method puts a premium on changes and offers inducements in favor 

 of instability, by an appeal to the vanity of would-be authorities, 

 which will not prove ineffectual. Some trinomialists disclaim re- 

 sponsibility for the repetition or duplication, and say they have 

 been betrayed by their disciples. 



The great aim of science being the diffusion of knowledge, it is 

 apparent that the more convenient, permanent and suggestive the 

 shape in which knowledge is put, the more easily it will be dif- 

 fused among men. In z-oolog} r the names are but symbols. Those 

 symbols are most perfect which are at once most convenient, most 

 permanent and most suggestive. Symbols of this character would 

 represent as far as possible our knowledge of the animals to which 

 they were applied. If inconvenient they would not be so easily 

 thought, spoken or written ; if not permanent to some degree they 

 might not continue in use long enough to receive general recogni- 

 tion, and thus they would introduce confusion and doubt as to 

 fact ; and if not suggestive they would compel the student to de- 

 pend on previous knowledge or subsequent tentative search for 

 clews to identity and history. 



Objections to the use of polynomials, including trinomials as 

 now used by some, are that they are inconveniently long, that they 

 are less likely to be permanent than binomials, that they do away 

 with certain advantages derived from long use of the latter, that 

 they increase the labor of tracing, and that they are less sugges- 

 tive as clews to history, etc. The binomials introduced by Linne, 

 though less suggestive in some respects, were so much more con- 

 venient that they at once supplanted Artedi's names in which con- 

 cise descriptions occupied the place of the specific title. Following 

 the polynomial idea to its logical sequence we should have bino- 

 mials for species not subdivided, trinomials for such as have va- 

 rieties, quadrinomials for such as have subvarieties, and so on. 

 One animal or plant would have a name of two words, and its 

 closely allied neighbor might have one of three, four or half a 

 dozen. To secure uniformity in the length of these names, repeti- 

 tion is the only alternative. A system of such possibilities is not 



