° 191.5^ ] Saunders, Recording Bird Songs. 179 



cult to distinguish between notes that are slurred together by 

 the bird and those that are clearly separate. This difficulty is 

 at best a slight one. The disconnected appearance of the songs 

 may seem great at first glance, but becomes insignificant as one 

 becomes accustomed to the method. 



By this same method it is also possible to represent the variations 

 in intensity of a bird song. This could be done by variation in the 

 breadth or heaviness of the lines, making heavy lines for the loud 

 notes, and light lines for the softer ones. I have not yet attempted 

 to measure the intensity of bird songs in the field so have omitted 

 this factor from the illustrations. 



The factor of pronunciation is one that presents some difficulties. 

 To just what extent birds produce recognizable vowel or consonant 

 sounds in their songs it is hard to say. It is probably true that a 

 purely musical note has no real vowel sound and that the only 

 difference in such notes is that of quality and not pronunciation. 

 Consonant sounds, however, may be occasionally recognized in bird 

 songs and call notes. The "k" sound in the call note of the crow, 

 for instance, is universally recognized. In true songs I believe that 

 the explosive consonants, such as " p," " k," "t" etc., are rare. The 

 commonest consonant sounds are liquid ones, such as "1" and 't", 

 connecting different notes. In the songs I have studied and 

 recorded, the liquid "1" is the only consonant I have recognized. 

 This sound is quite common in the songs of many species and is 

 evidently an important distinguishing character. I have repre- 

 sented the presence of this sound by a loop, at the beginning of the 

 note introduced by it, as shown in the songs of the Robin (fig. 6) 

 and Redwinged Blackbird (fig. 5). 



One of the first things that one notes after studying songs for a 

 time in the field is that even the simplest and commonest songs 

 are tremendously variable. This variation extends not only to 

 different individuals, but also to different songs by the same in- 

 dividual. The song of the Meadowlark is one that is quite simple 

 and easy to record, and yet shows enough variation to make a very 

 interesting study (fig. 7) . I have recorded thirty different songs of 

 the Meadowlark by the graphic method, and believe that with time 

 and opportunity I could record three or four times as many. Seven 

 of these songs were sung by the same bird during an hour's time. 



