KENNICOTT'S WILLOW- WARBLER 331 



and 61-67 in females. This places Ridgway's measurements within 

 the range of variation of the typical form. However, Dr. C. B. Tice- 

 hurst (1938), the leading and most recent authority on the group, 

 gives the measurement for males of the typical form as 65-72 milli- 

 meters and of females as 62.5-66 (exceptional specimens 58.5 and 61.5), 

 and as this is based on a large number of specimens it would seem as if 

 an average of 62 millimeters probably does indicate a smaller race. 

 Ticehurst himself was not able to examine sufficient material to judge 

 the validity of the Alaskan race, but quotes the opinion of J. L. Peters, 

 whom he consulted. Peters wrote: "This seems a rather unsatis- 

 factory race, but I think it may be distinguished from typical borealis 

 by the average smaller size and smaller bill. I compared 4 cf cf and 

 1 9 from Alaska with a series of Siberian breeding birds of borealis; 

 kennicotti cf wing 63.4-67.1, 9 63.1 mm. The colour characters 

 given by Ridgway do not seem to hold." 



Ticehurst adds that he has seen one unsexed bird from Alaska (wing 

 61.5), "which is much yellower below than autumn borealis, and has a 

 smaller bill than 9 borealis," thus just the reverse of Ridgway's de- 

 scription so far as the underparts are concerned. He further notes 

 that Friedmann (1937) also remarked on the yellowish wash on the 

 underparts of two birds from St. Lawrence Island, Alaska, taken in 

 July and August. Altogether one cannot but agree with Ticehurst's 

 final conclusion that "the form requires much further study on larger 

 material before its validity can be assured or its range demarcated." 



Swarth (1934) not only queried the validity of the race kennicotti 

 but even raised the question whether the species really breeds in Alaska 

 at all. He writes: "It is an extremely rare bird on the Alaskan side, it 

 has not been found actually nesting there, and occurrences are nearly 

 all as in our Nunivak specimen, of migrating birds in late summer. 

 These might be merely an overflow of migrants from the Siberian side 

 that later retrace their course." 



Reference to the original records for the localities of presumed 

 breeding quoted in the A.O.U. Check-list shows that Swarth's comment 

 was not unjustified. The earlier records for the coastal districts of 

 western and southwestern Alaska include only two June dates, namely, 

 June 14 (Grinnell, 1900) and June 19, the date of two specimens from 

 Nushagak in the National Museum, collected by J. W. Johnson in 

 1884 and mentioned by Osgood (1904). Osgood himself secured two 

 specimens near Iliamna at the base of the Alaska Peninsula on July 

 13 and 14, while none of the several records for St. Michael and the 

 vicinity is earlier than July 26 (Nelson, 1887). Most other records 

 refer to August or even September. No details are recorded about 

 the Nushagak specimens, and those secured by Osgood were asso- 

 ciating with other warblers, so there is little reason to think they were 



