316 Proceedings of Indiana Academy of Science 



other nest. This condition of clustered nests is due to the fact that 

 the choice of one individual in the same general strip of favorable terri- 

 tory is naturally the choice of many others. Possibly there also exists 

 a social instinct which causes the males to build nests in groups. These 

 nests were very close together and as many as 120 have been counted 

 in an area 150 feet in length and 15 feet wide. 



In late spring and early summer it is very easy to count the nests 

 since the water is very transparent although some of them were found 

 as deep as seven or eight or even ten feet. For the summer of 1921 



CD 



m 

 m 



^] 



Fi.tt. 1. Map of EhkIc" (Winona) Lake, showinR the number and distribution of 

 blut'jiill nests in 1921. The numbers indicate the location and number of nests in a t;i"oup. 



a total of 1,395 ne.sts were counted, 1,874 in 1922, and 2,398 for the 

 summer of 1923. 



To determine the average number of embryos per nest, four typical 

 nests were carefully collected and counted. The embryos were taken 

 from the nest immediately after they were hatched when most of them 

 still had the yolk sac attached. With the number of occupied nests and 

 an average of the individuals per nest, the approximate number of young 

 produced annually was determined. From the smaller nests the number 

 was obtained by counting them out individually, in the larger ne.sts 

 approximately 1,500 would be counted out, the water filtered off and 



