_Nepenthes_macfarlanei_ is correct.
The rule is:
If a non-latin name is latinized
If the name (of a person) is male
If the native form of the name (NF) ends with one of ("e", "i",
"j", "o", "u", "y", "er")
Then the latinized form of the name (LF) = NF + "us"
Else
If NF ends with "a"
Then LF = NF (NB: LF is still male!)
Else
Then LF = NF + "ius"
Else (i.e. name is female: I do not know the rules for neutral persons!)
If NF ends with one of ("e", "i", "j", "o", "u", "y", "er")
Then LF = NF + "a"
Else
If NF ends with "a"
Then LF = NF (LF female)
Else
Then LF = NF + "ia"
..
Genitive forms (GF) of LF:
If LF ends with "us"
Then GF = (LF - "us") + "i"
Else (i.e. LF ends with "a")
Then GF = LF + "e"
..
Adjective forms (AF) of LF:
If genus name (of the taxon) is male
If LF ends with "us"
If LF ends with "erus"
Then AF = (LF - "us") + "ianus"
Else
Then AF = (LF - "us") + "anus"
Else (i.e. LF ends with "a")
If LF ends with "era"
Then AF = (LF - "a") + "ianus"
Else
Then AF = (LF - "a") + "anus"
Else
If genus name is female
If LF ends with "us"
If LF ends with "erus"
Then AF = (LF - "us") + "iana"
Else
Then AF = (LF - "us") + "ana"
Else (i.e. LF ends with "a")
If LF ends with "era"
Then AF = (LF - "a") + "iana"
Else
Then AF = (LF - "a") + "ana"
Else (i.e. genus name is neuter)
If LF ends with "us"
If LF ends with "erus"
Then AF = (LF - "us") + "ianum"
Else
Then AF = (LF - "us") + "anum"
Else (i.e. LF ends with "a")
If LF ends with "era"
Then AF = (LF - "a") + "ianum"
Else
Then AF = (LF - "a") + "anum"
..
Examples:
Utricularia muelleri (GF after Mueller, m.)
U.nayarii (GF after Nayar, m.)
U.holtzei (GF after Holtze, m.)
Drosera hamiltonii (GF after Hamilton, m.)
D.burmannii (GF after Burmann, m.)
Nepenthes macfarlanei (GF after Macfarlane, m.)
N.eymae (GF after Eyma, m.)
N.burbidgeae (GF after Burbidge, f.)
N.hookerae (GF after Hooker, f.)
N.hookeri (GF after Hooker, m.)
N.hookeriana (AF after Hooker, m.)
N.northiana (AF after North, f.)
N.neufvilleana (AF after Neufville, m.)
U.buntingiana (AF after Bunting, m.)
U.glazioviana (AF after Glaziou, latinized "glaziovius", m.; well, just a
tiny exception to the rule...)
Latin terminations may be corrected (only!) if they were *wrong* in the
protologue (cf. also "Drosera pumila", originally spelled "pumilla". There
is no "pumillus" in Latin, only "pumilus" or "pusillus". In favour of
E.SANTOS let us assume she did not intend to create an illegitimate homonym
of D.pusilla H.B.K, so she must have meant "D.pumila"). Corrected names are
cited as if they were spelled correctly in the protologue (So it is
D.pumila SANTOS).
I hope the (still very primitive) rules above are sufficiently precise
(they may include errors!). Furthermore, there is almost no grammatical
rule without exceptions.
All other cases are (sometimes much) more complicated. Note that some names
(e.g. geographic names) are frequently not latinized non-latin names but
rather translations. In these cases all rules (i.e. an extended set...) of
Latin declension apply. Sometimes not a single person but a whole family is
commemorated (_Pinguicula_hintoniorum_ after family HINTON, formed via
latinization: "hintonius", and plural genitive: "hintoniorum").
Kind regards
Jan