Rafinesque
but Rafinesque 1815 steals the winners prize; most of his names are corruptive, based on earlier authors.
You must read his
Principes fondamentaux somiologie (1814) to start understanding (a bit) what he did. Rafinesque built his own, complex system of nomenclatural rules (that no one but him ever started to use, so far as I'm aware), which invalidated a significant number of names. In
Analyse (1815), he then replaced a considerable number of names in application to his rules, but indeed without explaining which rule(s) was/were causing the problem(s) -- he wrote that he would explain the changes he adopted there elsewhere (see [
p. 51]) but never did -- thus we can only guess, which in most cases is quite difficult due to the complexity of the system.
One of the names flagged as "no expl." in the
Key and cited from
Analyse actually dates from
Principes, and was explained there:
Anseria Rafinesque 1814 -
65.: 33. Règle. Les noms génèriques qui sont formés par la soustraction ou l'addition d'une ou plusieurs lettres ou syllabes au commencement ou à la fin d'un autre nom générique, ne peuvent pas être tolérés, il fant les changer ou en altérer la terminaison en telle sorte que le Genre radical devienne méconnaissable.
Obs. Cette règle se lie avec la précédente & elles se supportent réciproquement, les Genres Talpa L. & Catalpa J., Bromelia L. & Melia L., Cancer L. & Anser Brisson, Sinapis L. & Apis L. en sont des exemples; il faut dans tous les cas semblables conserver le nom antérieur (à moins qu'il ne soit d'ailleurs moins convenable) & modifier les autres; ainsi il faudra adopter Catalpium R., Ananas T., Anseria R. & Apicula R. au lieu de Catalpa, Bromelia, Anser & Apis.
Anseria Rafinesque 1814 was explicitly proposed to replace
Anser Brisson, which Rafinesque regarded as conflicting with (preoccupied by -- sort of) the crab genus name
Cancer Linnaeus 1758, because he did not admit generic names differing only in the presence/absence of one or more letter(s) at the start or end of the word.
One rule that seems to have been rather important to him was that, in his view, a generic name was ideally 3-4 syllables; 2- and 5-syllable names were admissible, but only if they did not break any other rule. (Even "soft" rules, that in isolation would have made the name "less good", but not not
per se unacceptable.) Monosyllabic names and less-than-perfect disyllabic names had to be lengthened, names with more than 5 syllables, and presumably less-than-perfect 5-syllable names, had to be shortened (even if this was at the cost of corrupting the stems).
In the
Key, a number of names that are replacement names in
Analyse, are apparently not readily interpreted as such. In Rafinesque's lists of animals in this work, each recognised taxonomic genus received a number. When two names were placed under the same number, these names were treated as synonyms -- the first name, often attributed to Rafinesque himself, being then intended to replace/be used instead of the second one, which was felt improper for some reason.
Thus, to stay with wildfowl,
Anassus Rafinesque 1815 -
Sous-famille. ANSERINIA. Les Ansèriens. Bec large, sans crêtes ni places nues sur la tête. G. 8. Anseria R. 9. Anassus R. Anas L. 10. Camploris R. 11. Solmorincus.
...is a replacement name for
Anas Linnaeus 1758, used here for the ninth genus of family "Serrirostria" (with
Anas cited as a synonym). Possibly (?) conflicting with
Ananas (accepted in
Principes;
Anas is senior, but shorter than ideal, which might have made the preference lean towards
Ananas). I'd suggest that
Anassus may be best interpreted as a simple, arbitrary, lengthening of
Anas. (And it is in any case not a synonym of Anser.)
Similarly,
Sularius Rafinesque 1815 -
2. Sous famille PLOTTIDIA. Les Plottidiens. Bec serreté. G. 4. Sularius R. Sula Lac. 5. Phaeton L. 6. Plottus L.
...is a replacement name for
Sula (attributed to Lacépède by Rafinesque), presumably also formed by arbitrary lengthening of the original disyllabic name -- which may also have conflicted with some other name, albeit I'm not clear which.
(There is no need of a question mark in "
?syn. Sula". Also, saying that
Sularius is "placed in Ptiopodia, sub-family Plottidia, before
Sula" as the Key currently does is a bit misleading, as this suggests that
Sula was treated as valid and distinct. It was neither.)