12 



(b) The four are distributed through two locks ; fourteen chances. 



(c) The worms are distributed in all three locks ; twelve chances. 

 If necessary the chances for larger numbers of worms can easily 



be calculated by the following rule. Each succeeding increase by one 

 worm requires the number of chances to be trebled 1, 3, 9, 27, 81, 273, 

 etc. In every case there is only one chance of all falling into one lock. 

 The following table makes the arrangement clear. 



TABLE VI. 



To find the figure for any place in the column all in two locks, add 

 one to the figure immediately above and multiply by two. 



To find the figure for any place in the column "all in three locks," 

 multiply the figure immediately above by three and add to this the 

 " all in two locks " figure corresponding to the number multiplied. 



Using the distribution outlined above, and the numbers of worms 

 to be distributed as found from Table I, the statement has been calcu- 

 lated which is tabulated on Table VII in the middle division. The 

 first division of the table showing the actually observed facts. 



Comparison of these two sets of figures shows : (a) that the number 

 of bolls damaged in one lock is excessive in the calculation as compared 

 to the observation, and (b) that the number calculated to be damaged 

 in three locks is deficient to about the same extent. 



The calculation demands more single locks damaged than ob- 

 served ; it is therefore obvious that in some cases and under certain 

 circumstances one worm is capable of doing damage in two locks. 

 This must be due to the worms wandering from one lock to the other, 

 examination of the material is not likely to give a reason, but it may be 

 when the insect is searching for a way out before pupation, when 

 insects are normally restless, or it may be wandering in search of more 

 palatable food. This is where animate nature differs from balls and 

 holes. Whatever may be the cause, it may be stated that the fact 

 that worms travel from lock to lock has actually been observed in 



