2 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 122 



reflect the unity behind the undertaking, and in the present paper 

 current results will be combined with previously unpublished data to 

 estabUsh several theoretical principles that have emerged in the course 

 of the work. 



Throughout the studies, the behavior patterns have been analyzed 

 in order to determine w^hich movements were stereotyped and, among 

 these, which conformed to the concept of fixed action patterns. Thus, 

 certain behavioral units meeting acceptable criteria could be treated 

 in the same manner as morphological structiu^es and compared from 

 one species to the next in the classical manner developed by Lorenz 

 (1957) and Tinbergen (1951). The ground rules for such comparative 

 studies have been reviewed recently by Wickler (1961) and need not 

 be reiterated here. In my studies, behavioral evolution was ap- 

 proached at the level of discrete behavioral units; such a description 

 of discrete behavioral patterns allowed the formulation of descriptions 

 of more complicated patterns consisting of predictable sequences of 

 discrete units in time. Further analyses of the frequency of occiu-- 

 rence of different stereotyped units of behavior performed in a social 

 context permitted a description of species-specific patterns of social 

 structiu"e. A knowledge of social structm'e permitted a partial answer 

 to the following question: Do organisms that evolve independently 

 in different geographical locations toward the same ecological niche 

 also evolve similar social organizations? If the answer were yes, then 

 it would be possible not only to predict a social system from a knowl- 

 edge of niche requirements but also to determine from comparative 

 studies what major environmental adaptations correlated with a given 

 social system. Thus, the comparison of social systems demanded a 

 description of the behavioral units and, in addition, the description of 

 the different forms of social organization by means of some consistent 

 experimental methodology (Eisenberg, 1964). Working in the 

 laboratory with small rodents was possible only when adequate field 

 data w^ere available to correct for any inadvertent misinterpretation 

 brought about by captivity. Finally, methods of quantification had 

 to be developed that permitted a valid comparison among difterent 

 species. After several years, it became obvious that simple criteria 

 such as the presence or absence of a given behavior pattern would not 

 suffice in determining relevant species differences. Indeed, it appeared 

 that, under the same stimulus conditions, the different species were all 

 capable of exhibiting almost the same discrete units of behavior and, 

 unless major morphological differences interfered, only the different 

 frequencies of occurrence served to separate species. Relative 

 differences rather than absolute differences in behavior became the 

 rule and led the present author to conclusions quite similar to those of 

 Leyhausen (1965). 



