NOTES ON NOMENCLATURE, III. 125 



faite depuis long temps, mime avant la publication des deux ouv- 

 rages anglais ou se trouvent de tres-bonnes figures de notre 

 oiseau.' It is most probable that when he wrote his description 

 he took it from an example of P. macrolopha, instead of from 

 the specimen figured which was in the Paris Museum, and 

 which he may not have seen for a long time, and was confound- 

 ed in his mind with the newly-figured P. macrolop/ia." 



Virtually therefore it comes to this : Temminck figured cas- 

 tanea, (that the type proves), but he described macrolopha. 

 Now this opens up a wide question. 



First — In fixing a name, are we to abide by an author's des- 

 cription or figure, when these refer to distinct species ? 



I say (where the figure is by an artist, and not by the author 

 of the name) most certainly by the description, which is the 

 name-giver's own work, while the figure is the work of another 

 person. 



Second. — Where a description is such that it does not agree 

 with the species, to which the name it defines is applied or pro- 

 posed to be applied, can the name be saved by referring to the 

 type and showing that, though the description was erroneous, yet 

 the name was really applied to the given species ? 

 I say no, except under two special conditions : — 

 1st. — That the name that would otherwise have been rejected, 

 has by long use acquired a scientific fixity : 



2nd. — That the name that would otherwise supersede it shall 

 have been assigned by the same authority who bestowed the 

 incorrectly-defined name. 



Unless both these conditions exist, no reference to a type 

 should be allowed to save a name founded on a description 

 which is distinctly at variance with the species to which the name 

 was (or is supposed to have been) intended to apply. 



Even for saving the name under these peculiar and neces- 

 sarily rare conditions, I have no warrant, and I would not for a 

 moment insist on them. I merely suggest them to brother orni- 

 thologists as a possible and not unreasonable or illogical re- 

 laxation of the general rule, which is, broadly, that names not 

 properly defined or indicated must be rejected. 



But though this is the Code rule, as a matter of fact even 

 English ornithologists seem now too generally to hold that no 

 matter, how erroneous a description is, you have only to hunt 

 out the type to establish the name. 



To my miud this is illogical. A and B both give names to 

 given species, and describe these, not only imperfectly, that is 

 nothing, but distinctly wrongly on a material point; why 

 should A's name be discarded and B's retained, simply because 

 while A's type has perished B's is still preserved ? 



