POSITION ON TREES. 329 



their way to suitable food plants, passed their transforma- 

 tions and thus established new colonies down stream. 



To determine the actual resistance of these caterpillars to 

 water, the following experiments were made in July, 1895. 

 July 12, four jars were filled with water, and at 9 a.m. five 

 gypsy moth caterpillars of the fifth molt were placed in 

 each. In jar No. 1, half of the caterpillars were alive the 

 third day; the remainder died July 17. In jar No. 2, one 

 caterpillar died the first and two the second day ; the other 

 two lived until the morning of the fifth day. In jar No. 3, 

 one died the first and the rest the fifth day. In jar No. 

 4, one died the first and one the second day, the rest living 

 until the morning of the fourth day. These results show that 

 the gypsy moth caterpillars are able to live in water for two or 

 three days, during which time, if they should fall into swiftly 

 moving streams, they would be carried a considerable dis- 

 tance, and in this manner some of the colonies found along 

 the shores of streams may have been established. 



Position of Caterpillars on Trees. 

 To ascertain the normal position of the caterpillars of the 

 gypsy moth on trees, a tight board fence was built around a 

 medium-sized apple tree upon which a number of caterpillars 

 were feeding. The fence was five feet in height, and en- 

 closed an area of eighteen by eighteen feet. At a height of 

 three feet from the ground a wide band of raupenleim was 

 placed around the fence on the inside, and at the base the 

 earth was banked up to the height of one foot, in order to 

 prevent the escape of the caterpillars. Careful observations 

 were made every morning by Mr. W. L. Tower, on the 

 location of the caterpillars and the places where they had 

 gone to rest after they stop})ed feeding. As the number of 

 caterpillars on the tree was not always the same, no tabu- 

 lated list of their positions could be made ; yet the relative 

 number found each day, in the same place and position, was 

 more constant than could be expected. By adding the num- 

 ber found in each location and the number found on the tree 

 every day, the percentage in each place was found, to be 

 fairly constant. 



