374 Correspondence. ae 
mens actually collected is so overwhelmingly in favor of the accuracy of 
this disposal of them. Therefore why should not the original observer 
make the same assumption? If we adopt Mr. Taverner’s plan for all 
sight records of trinomially named birds we see no possible alternative 
but to abandon the use of subspecies entirely. Another point to be con- 
sidered in this connection is the case of species which are very close to 
one another such as the Black-capped and Carolina Chickadees; Olive- 
backed and Gray-cheeked Thrushes; Louisiana and Northern Water- 
Thrushes; Western and Semipalmated Sandpipers, Common and Long- 
billed Dowitchers, etc. Some observers, under certain conditions, can 
distinguish most of these in life, but there are others who surely cannot. 
How can Mr. Taverner’s plan be applied to these? Trinomials we may 
remind him do not represent degrees of difference but the fact of inter- 
gradation, and there are certain subspecies which can be separated far 
more easily than can some species. If subspecies are abandoned as such 
many of them will have to be elevated to specific rank, as all of them 
are in Sharpe’s ‘Hand List’ and certain other works. 
‘J. D.’s criticism points out no definite policy and we are not clear 
whether he has Mr. Taverner’s plan in mind or whether he would reject 
‘sight’ records of this kind entirely. We can hardly suppose that he takes 
the latter view since we think that everyone will admit that we gain some- 
thing by recording the fact that Evening Grosbeaks of some kind visited us 
last winter even though we cannot say just which race each flock belonged to. 
Now we are not rejecting Mr. Taverner’s plan entirely. We think it is 
an excellent one in cases where a reasonable doubt exists as to the identity 
of the subspecies or in intermediate territory where two subspecies merge 
one into the other. In fact the plan has already been used in ‘The Auk’ 
but unfortunately it has caused misunderstanding, for the following 
reason. It often happens either from preference or accident, that the 
binomial form (Junco hyemalis, for instance) is used to indicate the eastern 
race of Junco instead of the more proper trinomial form (J. hyemalis hye- 
malis), as was the general custom prior to the last A. O. U. ‘Check-List’. 
It is thus not clear without further explanation whether the binomial name 
refers to the eastern race alone, or to this whole group of Juncos without 
indication of any individual race, as Mr. Taverner would use it. It would 
therefore seem clearer to adopt the plan used by Mr. Mousley (Auk, 1917, 
p. 215) in recording a brown-headed Chickadee, 1. e. “‘ Penthestes hudsonicus 
subsp.?”’? Mr. N. C. Brown on the same page adopts Mr. Taverner’s 
plan and writes ‘‘ Penthestes hudsonicus,”’ but has to add a statement that 
the form of the subspecies was not determined, in order to make it clear that 
he was not recording the true P. hudsonicus hudsonicus. In the ease of 
‘sight’ records of closely related species Mr. Taverner has no suggestion 
and we can apparently only take the word of the observer if he be reason- 
ably reliable, although even in the case of reliable persons there must always 
be a certain percentage of error in such cases. Long experience in com- 
piling migration records leads us to place far less reliance upon the average 
