176 Saunders, Recording Bird Songs. [April 



The following method has occurred to me by which the pitch 

 and duration of a song may be represented graphically, when once 

 it is determined. This method is to represent the song by lines on 

 ordinary coordinate paper, plotting the pitch along the ordinate, 

 that is in a vertical direction, and the duration along the abcissa 

 in a horizontal direction. In order to see whether this method 

 was practical, I tested it in the field during the spring of 1914, 

 recording 104 different songs and call-notes, representing 18 

 different species. The species included both birds with simple 

 songs, such as the Junco and Phoebe, and others with more com- 

 plicated songs such as the Song Sparrow (fig. 3) and Purple Finch 

 (fig. 4). It also included some bird sounds not properly classed as 

 songs, such as certain call-notes of the Flicker and the scream of 

 the Red-shouldered Hawk. 



Of course this method is not without its difficulties. It is 

 usually impossible, even with the simplest songs, to record them 

 after one hearing only, and with a long continued song, it is only 

 possible to catch and record phrases here and there. Such diffi- 

 culties, however, would be just as great, or even greater in using 

 musical notation. I have tried several times to record bird songs 

 by musical notation, and am certain that this graphic method is 

 much simpler, and much more easily used and mastered than is the 

 other. In the matter of pitch, one does not have to ascertain 

 whether the bird is singing in three flats, five sharps or something- 

 else. If the bird flats a little, or uses an interval not strictly a 

 fifth, seventh, or some other known to human music, this fact may 

 be shown and need not be modified to fit the human standard. 

 In the matter of time the same things are true. Notes need not 

 be reduced to quarters and eighths when they really have no such 

 definite relations to each other, but may be represented in their 

 actual true duration. In short, the method, like the bird's song 

 itself, is natural, and does not follow any fixed rule of either pitch 

 or time. 



The unit of measurement of pitch is of course the octave, but 

 this is not divided into eight parts as on the musical scale, but into 

 twelve parts, representing the twelve half-tones. Thus B and C, 

 and E and F are shown in their true relations, half a tone apart, 

 and not, as on the musical scale, spaced the same distance apart as 



