450 MELVILL: THE PRINCIPLES OF NOMENCLATURE. 
(Il). Pre-Linnzean authors, as not being binomial, not to be 
considered. 
(III, IV, V). The generic name, as given originally to an 
assemblage which might be subsequently divided, not to lapse, 
but to be allowed to be used for the most typical of the newly- 
constituted genera, ze, that species which the author con- 
sidered most typical. Should no type have been indicated 
by the author the original name to be kept for the subsequent 
sub-division that first received it. 
(VI, VII). A later name for the same genus to be reduced 
to a synonym, provided the genus be retained as one individual 
whole. If, however, it be subsequently divided by authors who 
have selected as types the sections now raised to the dignity 
of two genera, both the names to be kept. 
(VIII). A later name, however, embracing several earlier 
ones, not to stand by any pretence whatsoever. 
(IX). Should a genus be extended, so as to embrace one or 
two already defined and named, the earlier name to be chosen 
for the whole number of species. 
(X). No two genera to bear the same name. If by inadver- 
tence this has occurred, the older to stand. 
(XI, XII). A name, given under mistaken ideas, to be 
allowed to be changed, as also names not clearly defined or 
described. 
(XIII). When a specific name is adopted as a new generic 
term, the species, to avoid tautology, to be re-christened. 
(XIV). Latin terms and words to be used, or else Latinized 
Greek. Rules are given for Graeco-Latin terminology. 
In addition, the following recommendations were given for 
the improvement of nomenclature :— 
Genera to be substantives, not adjectival in their deriva- 
tion. 
Geographical names only to be recommended if the object 
be endemic there. 
J.C., vitt., Oct., 18976 
