12 
others for special approval or the re- 
verse. 
Rule 18. ‘If it be held that the generic 
and special names may not be tautonymic, 
the law of priority will determine whether 
the special or the generic name should be 
changed, e. g., Cossus cossus, L.’? The proviso 
in this case nullifies the rule for us, because 
we do not forbid tautonymity. Our prac- 
tice is bad—obviously so, on the score of 
literary propriety ; it is to be strongly dis- 
couraged ; but it seems an unabatable nui- 
sance, which most naturalists will put up 
with perforce, and it has one redeeming 
feature—we know absolutely what every 
tautonym means. 
Rule 19, bringing up ‘three classes’ of 
invalid names, is likely to remain a bone of 
contention in one-third of its scope. We 
all agree regarding homonymous names, 
that they are absolutely to be rejected; so, 
of course, regarding synonymous names. 
But not so regarding homophonous words, 
for surely no one of us would reject Seiwrus 
because its sound when spoken is identical 
with that of Sciwrus. 
Rules 20 and 21, hanging on the forego- 
ing, are open to difference of good sound 
opinion, possibly because, for one reason, it 
may not be always clear what words come 
under these provisions. It is desirable, but 
probably impossible, to have rigid rules 
here. No rule can possibly be more rigid 
or stable than the sum of the cases to 
which it applies; but the cases intended to 
be covered by Rules 20 and 21 are so shifty 
and mobile that no seive can be devised 
with meshes fine enough to catch them all. 
We confess ourselves puzzled here; we can- 
not offer better rules than Lord Walsing- 
ham has, yet we doubt their sufficiency. 
It seems to be a case where common sense, 
tact and expertness may work better than 
any formality. Let us agree, as most, if 
not all, ornithologists would, that Telea 
invalidates Teleia, and Pandemos invalidates 
SCIENCE. 
[N. S. Vou. VI. No. 131. 
Pandemis; does Ucetia invalidate Eusesia ? 
Personally, we should say that would de- 
pend upon the etymology in the case. If 
Ucetia be merely a bad way of spelling 
Eusesia, the etymon in the two cases being 
identical, we should say they were the 
same word, not available for two names, for 
all that they happen to be spelled so dif- 
ferently. We would, therefore, spell to the 
best of our ability both names as one, and 
use the right spelling or not according 
to the provisions of some other rule re- 
garding homonyms. But if Ucetia and 
Eusesia be of different etymons, only acci- 
dentally homophonous, we should regard 
them as distinct words, neither of which 
would invalidate the other. Yet we know 
that others might take exactly the opposite 
view and argue strongly in its favor. As 
we said in substance, this whole class of 
cases has thus far proven refractory to, or 
elusive of, any rule naturalists have had 
the wit or ingenuity to devise. 
Rule 22. ‘“‘ A name wrongly written is in- 
valid if, on legitimate correction, it becomes 
homonymous or homophonous with a 
valid name; e. g., Grapholitha, Hb., invali- 
dates Grapholita, Tr.” The A. O. U. Code 
would have simply to quash this rule, be- 
cause it forbids all correction of names, 
‘legitimate’ or illegitimate. We think it 
is the very worst blot on our Code, which 
has done more to bring that work into dis- 
repute than all our other weak or bad 
points put together. It puts a premium on 
ignorance or carelessness to rule, as we do 
in Canon XL., that a name shall endure 
exactly as it originally appeared in print, 
no matter how mangled it was, unless a 
typographical error be evident. Why be so 
fierce with the poor compositor, and let the 
guilty author go scot free? We need not 
become formal impurists for fear of purism. 
Personally, we decline to bind ourselves to 
misspelt words forever, for no better reason 
than that some zoologists, too ignorant or 
