136 Henshaw on the Hermit Thrush. 



Auduboni is rather better marked. I have never seen the wing of 

 paUasi four inches long, and doubt that it ever exceeds this dimen- 

 sion, as is the case with some examples of auduboni." 



Some of the statements contained in the above quotation were 

 especially interesting to me, as, after a large field experience, and 

 having collected many specimens of all three birds, I have never had 

 the slightest difficulty in discriminating between the forms. This 

 has doubtless been due in part to the fact, which the experi- 

 ence of every field-worker will attest, that various slight differences 

 of color, as well as certain other points, are perfectly apparent in 

 freshly killed specimens which are often partially or even wholly 

 lost in dried skins. 



But the main point involved is a matter of simple measurement, 

 namely, to ascertain whether the three forms do or do not intergrade 

 in size, and, if they do, to what extent ; in other words, to deter- 

 mine their relations by means of rule and dividers. For this pur- 

 pose Mr. Ridgvvay has kindly placed at my disposal the large num- 

 ber of specimens contained in the Smithsonian Institution, which, 

 together with the series collected by the Survey West of the 100th 

 Meridian, has supplied ample material, and I have been able to 

 include in my examination and to measure over 100 specimens* 

 divided as follows : of T. pallasi, 32 specimens ; T. auduboni, 39 

 specimens ; T. nanus, 35 specimens. 



The following figures represent the averages obtained : — 



T.pallasi, Wing, 3.61. Tail, 2.87. Bill, .53. Tarsus, 1.15. 

 T. auduboni, " 3.99. " 3.12. " .55. « 1.14. 



T. nanus, " 3.44. « 2.78. " .48. " 1.11. 



The number of specimens used would appear to be sufficient to 

 give a result closely approximate to correctness, although, as will be 

 evident from facts given further on, the greater the number used, 

 the wider will be the gaps between the three forms. 



From the above it will, I think, be at once apparent that the dif- 

 ferences here indicated express something more than mere individual 

 variation, and that the discrepancies in size, bo far as the average is 

 concerned, arc fully up to the requirements of varieties. The ex- 



* It is to be regretted thai the tables of measurements cannot 1»' here given 

 in full, but lack of. space Forbids. They will be presented elsewhere, and oppor- 

 tunity thus he afforded to compare the amount of individual variation, which, it 



may be Stated, is wry gnat. 



