population variability exclusively to the genetically fixed song. 

 This tendency is also clearly noticeable in the -work of Marler and 

 Tamura (1962) on the voice of Zonotrichia leucophrys . In our opinion, 

 the absence of uniformity in the local dialects of various species and 

 groupings is an additional argument in favor of the further development 

 of hierarchical classification of dialects, and their comparison with 

 the classifications of populations existing in ecology. The need for 

 further study is already definitely felt. 



Of special interest is the system worked out by N.P. Naumov, 

 according to which genetically isolated populations regularly merge 

 into a single system with phenotypical populations. The attempt to 

 apply the same principles to the classification of vocal dialects 

 appears justified to us, especially since with birds, as we have 

 attempted to explain, the voice happens to be the most flexible and 

 convenient mechanism for fixation (marking) of any grouping, beginning 

 with single specimens and pairs (as is the case with antiphonal species), 

 and up to several groups of families and populations. 



The principles and methods of coding geographic information in the 

 voice of animals present a somewhat separate problem. Valuable mat- 

 erial has accumulated on this subject in recent years. It happens 

 that geographical variability of the voice has a close, though rather 

 peculiar, relation to the apparatus of sound analysis. This circum- 

 stance has become especially important after the discovery of the so- 

 called "secondary" or "accessory" sounds in the voice of animals: 

 sounds which, according to some authors, have no functional signifi- 

 cance. These sounds arise together with the rendering of the basic, 

 biologically important ones. Thus, in the voice of birds there are 

 ultrasonic frequencies reaching k5, 000 cps, while it is currently be- 

 lieved that the upper limit of their sound registration is only 29,000 

 cps (Schwartzkopff, 1962). W. Thorpe and D. Griffin (1962), who made 

 a special investigation of this problem, have found ultra-sounds in the 

 voice of Locustella naevia, Erithacus rubecula , Ember iza calandra , 

 Sylvia communis , Acrocephalus scirpaceus , and others. Inasmuch as 

 naturally registered sounds have a functional significance, it is tempt- 

 ing to connect the nonadaptive geographical variability of voice with 

 secondary sounds. However, this assumption requires further investiga- 

 tion and development. Besides, the possibility is not excluded that 

 many of those sounds which until now have been considered as secondary 

 may in reality have a functional significance. 



The importance of the sound analyzer* as a system limiting any 

 functional reorganization in the voice is first of all expressed in the 

 extreme differences in acoustic means utilized for the transmission of 

 information by groups having different levels of hearing. 



* [According to Pavlov, an analyzer consists of the nerve endings, 

 peripheral nerves and cerebral connection of one particular faculty; 

 there are thus auditory, olfactory, gustatory, cutaneous and motor . ] 



