STUDYING THE HABITS OF FISHES. III7 



animal behavior. Every accurate and complete account of such behavior 

 is made up by putting together such answers. 



Repetition of observations. — Many observations are necessary on each point. 

 A single observation rarely suffices to answer any one of the many little ques- 

 tions that the observer puts to himself. The rapid movements of fishes in a 

 more or less turbid medium, the surface of which is rarely wholly smooth, make 

 observation difficult and increase the chances of error. The writer has in the 

 case of some difficult points repeated his observations a hundred or more times 

 before he has felt sure of their correctness. 



Importance of field experimentation. — Field observations should be checked, 

 wherever possible, by field experiments. The analysis of the problem, as ad- 

 vised in the third section above, sometimes brings to the front questions that 

 can not be answered by direct observation, so that recourse must be had to 

 experiment. Thus the building of the nest of Amia has never been observed, 

 for the probable reason, as the observations of the writer show, that the work 

 is done at night. In order to learn whether this work was done by the male 

 or female, the writer introduced males into a considerable bay access to which 

 was barred to the females by a net stretched across its mouth. By this device 

 he learned that the nests are built by the males. In work on the breeding 

 habits of Lampctra he has been able to make still more considerable use of the 

 experimental method. There is no doubt that the use of the method may 

 be extended to other forms. 



Making of records. — Records of all observations should be made on the 

 spot. It is not sufficient to observe what takes place and to write up notes in 

 the evening or at some more remote convenient time. The memory can not 

 be trusted to retain accurately the details of happenings so complicated as those 

 that fall under the eyes of the observer of living fish. It would seem hardly 

 necessary to insist on a precaution so obvious were it not that the observer of 

 fish is placed under sore temptation to defer the writing of his notes. The 

 scenes that he has before him are of absorbing interest and require his closest 

 attention to follow them and unravel their complications. He is loth to spare 

 any time to note taking. And not only does the taking of notes consume time ; 

 it causes the loss of observations, at least for the time being; for while the observer 

 is recording one occurrence another follows it, and is lost to him. This loss can 

 only be made good by repeated observation, and while the opportunity for such 

 repetition may be long in coming, it is better to have an accurate record of a 

 part of what has happened than to have an inaccurate record of the whole. 

 Therefore, detailed notes should be made on the spot. There is another reason 

 for this procedure, namely, that the writing down of an observation forces the 

 conscientious observer to be accurate, and shows him wherein his observations 

 are incomplete. A field note, once written, suggests some other query, so that 



