How the Green Heron Wandered 203 



the Pacific Coast. One of them, the Frazar green heron, was 

 restricted to the peninsula of Lower California and could be 

 eliminated. The other, the Anthony green heron, had been 

 found in 1895 by an ornithologist named Mearns and de- 

 scribed by him from specimens taken in the Colorado Valley 

 near the Mexican boundary. It was a lighter-colored bird 

 than its eastern relative, and at first was supposed to occur 

 only in the arid regions of California and the Southwest. 

 There it breeds in trees along the irrigation ditches and 

 streams. I found that, by 1926, the range of this supposed 

 specialist in arid localities had been extended as far north 

 as Fort Klamath in the southern part of Oregon, and that 

 one person had reported the bird to be a "not common resi- 

 dent" in Washington. The fact that the best of the north- 

 western ornithologists— and there were many at that time- 

 made no report of the occurrence of this bird in Washington 

 until many years after makes this record doubtful. 



Although no green heron had been reported as far north 

 as Seattle, I made no attempt to publicize this extension of 

 range. It would not have been accepted as a scientific record, 

 for there is an unwritten rule that such records should be 

 substantiated by the actual skin of the bird. This is probably 

 sound from the viewpoint of science. It aids greatly in keep- 

 ing the records free from errors and frauds due to ignorance 

 or evil intention, although it would not prevent an unscrupu- 

 lous person from collecting skins in one district and then 

 representing them as taken in another. I quite agreed with 

 those who held that the indiscriminate acceptance of sight 

 records could be responsible for many errors. Had I wished 

 to claim a record I could easily have shot the visitor, but I 

 did not think that the matter of proving range extension was 

 of enough importance to justify killing the bird that had set 

 up the record. Then I thought of testing the value of color 

 photographs in the identification of species. Would it be 

 satisfactory? I decided to find out. 



