191(5] Introduction xiii 



student uf animal life slioulil he familiar with the objects of his interest in 

 nature as well as in the laboratory: that he should possess, as a basis for 

 evaluating the results of exiieriments, intimate knowledge of the instincts, 

 habits, temperaments, and habitat of whatever type of organism he happens 

 to be using for experimental purposes. 



Even if the natural environment could be duplicated, another 

 insuperable difficulty confronts the laboratory experimentalist. He 

 is compelled to restrict his experiments to a few individuals; btit 

 he always applies hi.s conclusions to races, varieties, species, or some 

 other similar group of organisms. How can he be sure that the behavior 

 of the individuals selected is typical of that of the group as a whole? 

 He cannot, no matter how large the number of individuals experi- 

 mented upon, unless he knows in considerable detail the environ- 

 mental conditions at the time and place his selected individuals were 

 collected ; and unless he knows how the group as a whole is distributed 

 with respect to variations in the environmental element he deals with 

 in his experiments. An illustration will make this clear. 



The species of chaetognatha most characteristic of the San Diego 

 region, Sagitta bipunctata, increases in abundance on the .surface from 

 noon until some time after sunset, when it attains a maximum ; but 

 large numbers may occur on the surface at all times of day and night 

 — even at noon. Is it reasonable to assume that individuals collected 

 at noon would react in the same way to variations in light intensity 

 as tho.se collected after sundown? Yet the latter are most typical of 

 the .species as a whole. Again, the species decreases in abundance on the 

 surface as the temperature increases above 15° C, but large numbers are 

 normally present at all temperatures. Is it reasonable to assume that 

 those collected on the .surface when the temperature was 20° C would 

 react to variations in temperature like those collected when the tem- 

 perature was 15° C? Yet the latter are most typical of the species as 

 a whole. Similar phenomena occur with respect to variations in 

 salinity, and probabl.y with respect to variations in gas-content, and 

 other environmental elements. 



Obviously, the magnitudes of each element of an environmental 

 complex may act like so many .sieves separating the individuals of any 

 one species into a variety of behavior-classes. Reactions witnessed in 

 a laboratory, therefore, may be largely due to the particular individuals 

 collected. If this is so, intensive investigation of the organism in its 

 natural environment is a necessary preliminary to generalizations from 

 laboratory experiments on behavior : if it is not so, that fact must be 

 proved in each particular case, or conclusions applied to the species (or 



