PRINCIPLES, CANONS, AND RECOMMENDATIONS. 65 



modifying a Latin generic name, to indicate that a new genus thus named is 

 in some way related to the one whose name is thus modified. They are also 

 used in reforming a name which is inadmissible for any reason, in order to 

 preserve a suggestive and convenient similarity. For instance, Ccecilia, if 

 employed for a shell, but which was found to be preoccupied in some other 

 class, might be modified to Ccecilianella, in order that convenience in con- 

 sulting indices might be conserved for the new name in connection with the 

 old one. (DALL, Rep., p. 30.) 



17. Geographical specific names are formed by adding the suffixes -us, 

 -ins, -tcus, -inus, -itus, (or their feminine or neuter equivalents, as the case 

 may require,) and -etisis, the name itself suffering no modification except in 

 its termination. 



18. Manuscript names used by collectors in their notes or on labels, if 

 well chosen, may be adopted, the adopter of the name of course supplying 

 a description ; and he should further state that the name has not previously 

 been formally introduced. Without this precaution the use of manuscript 

 names is highly objectionable, and has been the source of great confusion 

 and annoyance. The manuscript names of Beck, Solander, Leach, and 

 others, have long been stumbling-blocks, from having been quoted by natu- 

 ralists with no reference to the fact that they were unaccompanied by descrip- 

 tions, and therefore without standing. (DALL, Rep., p. 33.) 



19. In subdividing an old genus it would be better to make the subdi- 

 visions agree in gender with that of the original group, in order that specific 

 names may be preserved unaltered. 



13. Of the Transliteration of Names. 



RECOMMENDATION IV. Names adopted from languages writ- 

 ten in other than Roman characters, as the Greek, Russian, 

 Arabic, Japanese, etc., or from languages containing characters 

 not represented in the Roman alphabet, as the Spanish, French, 

 German, Scandinavian, Western Slavonian, etc., should be ren- 

 dered by the corresponding Roman letters or combinations of 

 letters. 



REMARKS. The transliteration of letters not Roman into those of the 

 Latin alphabet is a matter of some difficulty and uncertainty, as philologists 

 are not yet in agreement as to the rules. The only alphabet in regard to 

 which scholars nearly agree' being the Greek one, the commonly adopted 

 system should be followed, and also in case of names derived from the 

 modern Greek language. In regard to the other alphabets, it is to be 

 recommended that in transliterating the spelling be as nearly phonetic as 



5 



