44 
thp: condor 
1 VOL. V 
Tlie writer is speaking solely for himself, and no matter whether his views be 
ultimately endorsed or condemned by the A. O. U. Committee, he wishes to be 
understood as always advocating adherence to the Code and Check-list of the A. 
O. U. as the only way to secure uniformity — the main object for which we strive. 
He is also well aware much the same problems as here presented have already 
been discussed by Dr. Merriam, Dr. Allen and others, so that no originality is 
claimed, but merely an expression of opinion. 
The use of trinomials, as established by the A. O. U. Code and adopted by 
most writers on vertebrates, is generally understood to be the designation of the 
geographic variants of a wide ranging form, which merge into one another where 
their ranges join- — ^i. e., incipient species, produced by peculiar environments, but 
which are not yet entirely isolated from one another. 'I'he forms .so designated 
are called subspecies. 
As a convenient method of deciding whether a certain race or form should be 
regarded as a species or subspecies, actual intergradation between contiguous forms 
was adopted by the A. O. U. Code as the criterion. 
When we come to name resident birds of coastwise islands which are but 
slightly differentiated from the mainland stock we at once confront a problem. 
Intergradation in the sense of interbreeding is impossible, consequently some 
writers maintain that all island forms must be regarded as species (binomial). But 
intergradation in the sense of overlapping of characters exists in many cases, and 
on this ground others term them subspecies (trinomial) and in practice a form is 
judged to be a species or subspecies by the degree of difference exhibited between 
it and its nearest geographic relative. This latter would seem to be the more log- 
ical course, since by the former plan we might ju.st as well separate the song spar- 
rows for instance of San Clemente and San Miguel Islands since they are geo- 
graphically separated into two races, though as yet we cannot detect any tangible 
difference between them! 
Extending the practice of recognizing overlapping of characters as intergrada- 
tion, we find the geographically isolated though closely allied Florida burrowing 
owl listed as a subspecies of the bird of the plains, and other similar cases culmin- 
ating with Mr. Nelson’s recently described “subspecies” of the Cuban cliff swal- 
low from western Mexico! This practice is sev^erely criticised by .some, but if we 
regard these forms as species solely on account of geographic isolation, what are 
we going to do with the martin of Southern Mexico which Mr. Nelson states is in- 
distinguishable from the Cuban martin. Surely we cannot separate it purely on 
geographic grounds and if we do not separate it we are calling by the same name 
two forms which have probably developed independently, and thus loosing sight 
in our nomenclature of a fact of evolution, the indication ot which facts is, accord- 
ing to the strict adherents of the actual intergradation principle, the main object 
of trinomial nomenclature. The question naturally occurs can we indicate in our 
nomenclature all these facts of evolution without seriously impairing the utility of 
our names as names? 
The foregoing cases are those in which actual intergradation is either impossi- 
ble or doubtful. Let us now consider some where it is admittedly a fact. 
Mr. Ridgway has shown that all of our continental .song sparrows pass im- 
perceptibly one into the other where their ranges touch, and as a result we have 
the little speckled-breasted bird of the California salt-marshes listed as a subspecies 
of the big gray bird of xAlaska, more than twice its size, and which any novice 
would regard as a perfectly distinct kind of bird. 
