INLAND FISHERIES. 31 



This tabulated statement does not require particular explana- 

 tion. By g-lancino- over the results of tlie various experiments 

 made at different times, a g-eneral idea will be obtained of the 

 proportion of clams which maj^ be expected to burrow when they 

 are sown on the surface. 



On September 13, 14, and 15 a new series of experiments were 

 made on the burrowing habits, in a more systematic way than had 

 been done heretofore. The clams selected for these experiments 

 were separated into four classes which are given successively in 

 the tables, viz. : 



I. Laro-e, 50-70mm. (2-2f inches) in length ; ] 



" ' / Set of 1899 or 



11. Medium, 35-50mm. ; - previous to 



III. Small, 25-35mm. ; ) **^^''^^*"- 



IV. Small set of 1900, 20-30mm. 



To insure the protection of the clams Avliile exposed, and to 

 prevent them from washing away, boxes with 8-inch sides and ^-inch 

 wire mesh bottoms were inverted over the beds. All the clams 

 were soAved on the surface just as the tide was beginning to rise. 



The attempt was made to determine the value of the following 

 conditions for each class of clams : 



a. Sowing clams immediately after digging. 



kept 24 hours out of water (dry). 



" " in sea- water, 



in soil softened by previous digging. 



" not dug up. 

 in gravel, 

 in sand. 



For convenience in referring to the tables, these letters are 

 placed in the first column. 



