NOMENCLATURE. 349 



prominent or peculiar character or appearance, from localities, 

 from the names of persons (especially of discoverers), from 

 indigenous or vulgar names, or even from arbitraiy combina- 

 tions of letters. Unmeaning names, if not in principle the best, 

 are never misleading. The main requisite is that they should 

 be euphonious, not too long, and that the}' should be adaptable 

 to the Latin tongue. Characteristic names, when possible, are 

 among the best ; such as Sanguinaria for an herb with red juice, 

 Htzmatoxylon for the Logwood tree, Lithospermum for a plant 

 with stony seeds (or seeming seeds) , Myosurus for a plant with 

 gyncecium resembling the tail of a mouse. Names of this sort 

 do not ftfw*ya hold out well ; for Chrysanthemum, so called from 

 its golden yellow blossoms, now has many white-flowered species, 

 Poly gala is wholly destitute of milk, and many species of Con- 

 volvulus do not twine. Neat anagrams are not bad, such as 

 Brown's Tellima for a genus nearly related to Mitella. Personal 

 generic names are wholly proper when dedicated to botanists, 

 especially to the discoverer of the plant, or to other naturalists, 

 or to persons who have furthered botanical investigation or 

 exploration. Ancient names of this kind have been mentioned, 

 also some of those which commemorate the earlier botanists. 

 (702.) At present, almost eveiy devotee of the science is thus 

 commemorated, from Linnaeus and Jussieu downward. In 

 forming such names, the name of the person, cleared of titles 

 and accessory particles (thus Candollea, not DecandoUea) , takes 

 the final -a or -ia and becomes feminine ; and its orthography is 

 preserved as far as possible, making only necessary concessions 

 to euphon}' and to the genius of the Latin language. 1 



The Linnaean canon forbade the use of the same generic name in botany 

 and zoology, a rule now impossible to maintain. Perhaps we cannot pre- 

 vent the duplication of phsenogamous names in the lower Cryptogamia. 



1 Thus, we may write Lescuria instead of Lesquereuxia, although Michauxia 

 is the form for the genus dedicated to Michaux, however pronounced. The 

 genus dedicated to Strangways is written Stranvcesia (although Strangwaysta 

 might have been tolerable) ; to Andrzeiowsky, Andreoskia ; to Leeuwenhoek, 

 Levenhookia (although the elder DeCandolle restored all the vowels), &c. 

 As specimens of overdone simplification, there is Gundelia, named for Gun- 

 delsheimer, and Goodenla, named for Bishop Goodenough, although Gundels- 

 heimera would not in these days be objected to, and Goodenovia is faultless. 

 Yet the names having been so introduced into the science should remain, 

 fixity being of more importance than perfection. Mistaken orthography 

 of the name itself may, however, be set right. Brown's Lecheruntltia is Les- 

 chenaultia, Nuttall's Wisteria (named after Dr. Wistar) is Wistaria. The 

 rule laid down in the code as drawn up by Alphonse DeCandolle is: 

 " When a name is drawn from a modern language, it is to be maintained 

 just as it was made, even in the case of the spelling having been misunder- 



