42 ANIMAL LIFE AND SOCIAL GROWTH 



subjected to toleration tests to determine the lim- 

 its of their resistance to the different elements of 

 their environment such as determining the heat 

 or cold that will just kill them, the amount of 

 humidity they can stand, the length of food or 

 water starvation periods before death, and so on. 

 All these experiments will be carried on in care- 

 fully cleaned laboratory rooms and containers 

 as well as under conditions roughly approaching 

 those normally found in nature. 



Whether by these quadrat methods or by some 

 other means of quantitative sampling, we can 

 readily collect an impressive amount of mathe- 

 matical data. These must be tested by statistical 

 methods to determine the coefficient of correla- 

 tion in distribution in order to find whether we 

 have unearthed a real relationship. At about 

 this stage in the analysis of the ecological relations 

 within the university or community, it may be 

 necessary to use other collecting methods. For 

 this purpose we may devise nets of known size 

 and mesh and test their efficiency in collecting 

 by dragging them through a measured amount of 

 space in which we have released a hundred fresh- 

 men or other easily recognized members of the 

 college community. With tree-dwelling animals, 

 if they are present, we may spread a canvas of 

 known size and beat or shake the tree in as nearly a 

 standard manner as is possible under the cir- 



