NATURAL HISTORY OF AMIA CALVA LINNAEUS. 91 



was much interrupted and broken up into irregular channels here shallow, there deep. 

 Nor is the course over which the swarm moves the only one open to it. There are 

 many other channels through which it could pass quite as readily as through the 

 one followed, and many of these lead to deep water quite as directly as the one fol- 

 lowed. In short, I could see nothing in the physical environment which could 

 determine the course of the larvae. 



"The whole movement of the swarm when leaving the nest suggests that it 

 follows the trail of the male by scent. The digressions of the swarm from the main 

 course of the male are probably made in following out digressions made by him 

 on his inward trip. These were particularly numerous at the beginning of his 

 inward course, and here the swarm became most spread out and most confused. 

 The swarm appeared here to follow out side lines of scent, and, finding them to end 

 blindly, to return always to the main line and continue along it. The behavior of 

 the swarm is much like that of a dog trailing a covey of quail." (Note-book.) 



The following experiment indicates that this is the real explanation. A rough 

 model of an adult Amia was made and covered with black rubber cloth. A freshly 

 killed male fish was then trussed up and attached to the end of a stick so that he 

 could be readily moved about in the water in his natural position and in a natural 

 manner. The male was then at once taken to a nest on which the larvae were 

 swimming as though ready to leave the nest, and was made to pass several times 

 over a course leading from the nest to a distance of about a metre. The larvse soon 

 followed this course and assembled about the male and could then be readily led 

 ahead for a short distance by slowly moving the male fish. When the model was 

 substituted they paid no attention to it. It was noticed, however, that they did 

 gather in numbers in the deep water near the heel of my rubber boot, and that after 

 a time the dead fish no longer attracted them, perhaps because the odor of the male 

 had changed. There was no opportunity to repeat the experiment. 



B. The larvae outside the nest. — (12 to 100 millimetres.) During their 

 life in swarms or schools, as I shall now call them since the individual and swarm 

 progressive movements are fully developed, the following changes take place in the 

 appearance and behavior of the larvae. All measurements of larvse are from the tip 

 of the snout to the tip of the tail. 



a. Color changes. — Larvae of 12 to 30 millimetres. At 12 millimetres the 

 colors are as already described. As the animals grow older the black of the back and 

 sides takes on a more greenish hue. 



Larvae of 30 to 40 millimetres. Larvae of 30 to 32 millimetres when taken 

 from the natural waters are colored as follows: very black above and on sides; 



