50 BULLETIN 475, TJ. S. DEPARTMENT OF AGRICULTURE. 
they are wonderfully diligent and expert in searching it out. Many 
experiments in the direct seeding of hardwoods and conifers have 
failed because nearly all of the seed was eaten by mice, chiptnunks, 
and other animals. Where tree seed attractive to rodents is to be 
sown it is necessary, therefore, to consider the systematic poisoning 
of the area as an essential operation. In numerous experiments con- 
ducted by the Forest Service, seed spots have been sown and then 
covered with small portable screens which effectively ^protected them 
from rodents and birds. The successful germination in these spots 
and the absolute failure in similar unprotected spots adjacent showed 
how important an adverse factor animals, and possibly birds, are in 
reforestation. 
In the spring of 1910, in cooperation with the Bureau of Biological 
Survey, intensive studies of the damage from rodents were conducted 
at a number of places where direct seeding was in progress, and on 
many other sowing areas observations were made by forest officers. 
In nearl}^ ever}^ case a different species of animal was found to be the 
chief cause of damage. There are, of course, a great many species 
of mice, chipmunks, and ground squirrels, and a number of different 
kinds of tree squirrels. Each species appears to have distinctive 
habits, and the food and activities of the same species differ greatly 
with the time of year. The kind of poison bait and the manner of 
applying it must therefore be adapted not only to the species of 
animal to be destroyed, but to its taste and activities at the time the 
poisoning is done. The latest approved methods of poisoning 
rodents shown below are those developed by the United States 
Biological Survey. 
The poison preparation which most nearly meets all requirements 
is as follows: 
Mix 1 heaping tablespoonful of gloss starcli in I teacup of cold water and 
stir with 1 pint of boiling water to make a thin, clear mucilage. Remove from 
the stove. Mix together 1 ounce of powdered strychnine (alkaloid)* and 1 
ounce of powdered bicarbonate of soda and stir with the starch to a smooth, 
creamy mass. Stir in 1 tablespoonful of glycerine, and finally J ounce of sac- 
charine. Apply to 20 quarts of good clean oats or wheat and mix thoroughly to 
coat each kernel. 
This poison is effective in destroying chipmunks, kangaroo rats, 
pocket mice, the smaller species of ground squirrels, and, at times, 
kills many white- footed mice. 
Oats are generally the most successful bait. On account of the 
skill of chipmunks in "hulling," wheat is more effective for these 
animals. Barley, in the proportion of 16 quarts to each ounce of 
1 If strychnine sulphate is used, dissolve in the boiling water before adding the starch ; 
also dissolve the bicarbonate of soda before adding to the poisoned starch. Care Is 
necessary that the resulting ml:sture does not greatly exceed a pint. 
