B 
Parasitization percentages for the other collections may be de- 
termined in the same manner. 
In many cases, as in the above example, it may not be con- 
sidered necessary to record the percentages beyond the tenths place, 
but the percentage tables have been carried to the nearest thousandth 
for problems in which this degree of accuracy is desired. 
The rounding procedure followed in these tables has been 
to add 1 if the following digit is more than 5, to disregard the 
following digit if less than 5, to add 1 if the following digit is 
an even 5 and the digit to be rounded is odd, and to disregard the 
5 if the digit to be rounded is even. 
Often it will be found that the range of percentage calcu- 
lations needed for a particular problem will be limited. Thus over 
99 percent of the hundreds of collections of corn-borer larvae made 
annually to determine parasitization fall in the range 80 to 120 
hosts collected, and parasitization, usually by several species, 
totals less than 50 percent. The thousands of percentage figures 
needed in this problem can all be read directly from only four 
pages (12, 14, 16, and 19) of the tables. 
For more convenient use it is suggested that the staples be 
removed and the tables bound in a good looseleaf notebook cover 
v/ith an ample series of fasteners. They can then be laid open flat 
on the desk at the pages particularly needed. 
In setting up tables of this nature errors occur in spite of 
very careful checking. Every effort has been made to make the 
present tables accurate, but it is possible that a few errors occur. 
It will be appreciated if anyone finding an error in these tables 
will kindly inform the author. The author would also be interested 
in learning if similar tables for numbers above 250 would be useful 
to research workers. 
