DETERMINATION OF GENERIC TYPES, ETC. 15 



Example. Psarocolius Wag}., 1827, is equivalent to five or six genera previously 

 published under other names, therefore Psarocolius should be canceled. 



If these previously published genera be separately adopted (as is the case with the 

 equivalents of Psarocolius) , their original names will of course prevail; but if we fol- 

 low the later author, in combining them into one, the following rule is necessary: 



A genus compounded of two or more previously proposed genera whose characters are now 

 deemed insufficient should retain the name of one of them. It sometimes happens that 

 the progress of science requires two or more genera, founded on insufficient or erro- 

 neous characters, to be combined together into one. In such cases the law of priority 

 forbids us to cancel all the original names and impose a new one on this compound 

 genus. We must therefore select some one species as a type or example, and give 

 the generic name which it formerly bore to the whole group now formed. If these 

 original generic names differ in date, the oldest one should be the one adopted. 



9. In compounding a genus out of several smaller ones, the earliest of them, if 

 otherwise unobjectionable, should be selected and its former generic name be 

 extended over the new genus so compounded. 



Example. The genera Accentor and Prunella of Vieillot, not being considered suf- 

 ficiently distinct in character, are now united under the general name Accentor, that 

 being the earliest. 



It will thus be seen that the principle of "page precedence" was 

 recognized by this Code, 1, but not as an ironclad law; the principle 

 of type by tautonymy also seems to be referred to, 1; further, the 

 principle of the first reviser is clearly referred to under 5; the prin- 

 ciple of '"type by inclusion'' is evident in 6. 



The principle of " type by tautonymy," apparently indicated in the 

 B. A. Code, is said to have first been advocated by Newton (1871, 1876, 

 1879). It was formulated by Carus and IStiles in 1898, and has recently 

 (1902) been formally adopted by a number of American zoologists. 



THE DALL CODE, 1877. 



In the Dall (1877a, 39-10) Code the following paragraphs refer 

 directly or indirectly to generic types: 



LI. When a group or genus is divided into two or more groups the original name 

 must be preserved and given to one of the principal divisions. The division includ- 

 ing the typical species of the primitive genus, if any type had been specified, or the 

 oldest, best known, or most characteristic of the species originally included when 

 the primitive genus was first described by its author, is the portion for which the 

 original name is to be preserved. If there is no section special!}' so distinguished, 

 that which retains the larger number of species should retain the old name (D. C.), 

 but the latter can not be applied to a restricted group containing none of the species 

 referred to the primitive group by its author at the time when it was described or 

 when he enumerated the species contained in it. 



The majority of the replies to query XII of the circular concur in the above. 



According to Limueus the name should remain with the most common and offici- 

 nal species; an equivocal expression if there is one which is most common and 

 another the officinal species. The Convolvulus sepiurn and the Erica rntgnris were 

 very common and very anciently named species when Brown made of one the genus 

 Calystegia, and De Candolle of the other, his genus Calluna. It was, however, much 

 better to. do this than to change the names of a hundred species of Convolvulus and 200 

 of Erica. When there is no authoritative type the number of species should always 

 be taken into consideration. (D. C. ) 



