19i()] Iiitrotliictloii XV 



fiiimarcJiicKs does not. Why? ()l)viously. no physii-o-elu'iiiical theory 

 like Ostwald's can explain siieh facts unless, as is not infrequently the 

 case, one selects only the facts which fit and ignores those which do 

 not fit. "Every truly vital chemico-physical problem of organisms," 

 says Ritter (1915, p. 231), "is two-phased: how do the chemico- 

 physical attributes of the constituent substances act upon and .so 

 explain the organisms; and what particular structures and activities 

 are the chemical substances caused to manifest by being constituents 

 of and used by the particular organisms?" Specificities in the vertical 

 migrations of S. bipvnctata and C. finmarchicus can be physico- 

 chemically accounted for only by corresponding specificities in the 

 physico-chemical makeup of the organisms. The main value of 

 Ostwald's theory consists in pointing out that variations in viscositj' 

 are involved in the phenomena of vertical migration. It. of course, 

 contributes something to the explanation of the.se phenomena, but 

 very little. 



Laboratory experiment and field observation must go hand in hand. 

 The former cannot, except by inference, ascertain the manner in which 

 a species is related to its environmental complex. The latter cannot, 

 except by inference, ascertain the nature of response involved in 

 correlations observed between marine organisms (or any other kind of 

 organisms) and their environments. Observation alone cannot deter- 

 mine whether the observed correlations are due to tropisms. trials and 

 errors, or some indirect metabolic reactions. Experiment alone cannot 

 reveal the fact that Sagitta bipunctata, for example, is usually more 

 abundant between fifteen and thirty fathoms than at any other depth ; 

 that it decreases in abundance below this depth nrach mor(> slowly than 

 it does above it ; that it maintains its maxinunn abundance at liigher 

 levels during the summer (July to October) than during the winter 

 (November to March) ; that it decreases in abundance as the distance 

 from the coast increases at all depths above twenty fathoms, while it 

 increases in abundance at all depths below thirty fathoms, etc. These 

 facts pertain to the species' behavior and have played their roles in its 

 evolution just as certainly and to fully as great an extent as is the ease 

 with any facts of behavior demonstrated in a laboratory. Assuredly, 

 both sorts of investigation are required in order to approach, even 

 remotely, complete knowledge of the behavior of any species. "That 

 old and true method of natural history," says Whitman (1902, p. 505), 

 " — observation — must ever have a large share in the study of living 

 things. Observation, experiment, and reflection are three in one. 



