116 
PROCEEDINGS OF THE NATURAL [April, 1887. 
bright red, 2 dark red and 5 blue—11 in all. On August 13 
among 227 killed we found 4 marked green, 4 blue and 1 red—9 
in all. Out of 384 wasps, then, 20 were marked, or about 1 in 
every 19. The marked wasps in the last lot that we killed had 
color nearly eaten off so that we did not take any more. 
^ e painted some of these wasps by putting a long slender 
brush through the door of the cage and touching them as they 
clung to the sides ; but as they immediately began to clean each 
ot er off we changed our plan and painted them as they passed 
out. 
From the results of our killing and counting we concluded 
tat ough the wasps started from the cage in wrong direc¬ 
tions, they, after a time, succeeded in reaching home. All of 
they wasps marked green started wrong and yet we found 4 out 
° • 0 wasps out of the 75 that we marked, started directly 
toward the nest upon being liberated, but even if 4 of those 
marked green had been among them we should have thought it 
extremely improbable that these were identical with the 20 
we afterwards found marked in the nest. However, any doubts 
which we might have had about it would have been settled by 
our other experiments. 
10 , ™ ‘°° k 20 ”»»ps from nest nnmberone, which 
:~z , v woods on n piateau ' °" e ■»>««! 
TmUe Z. ,h Z' “ d them frM °" water • q— 
bo» sZ Z FiVe etarted off b « soon returned to the 
second Z Z m* “ eW aWOy a 8 ai ”. * 
in different directions ' Z's lhen f°° k ‘ he ‘ r , “ l d ” p * rtnr ‘ 
;:~p* 7 k 
On the evening of August 18 J n ! fromlt - 
we closed the entrance to the nest no thl " 
nor ont. Early on the mor ■ 80 tha ‘ none could pm m 
There were no wasps to 19th We went to the ne8i 
a few came straggling home and th* 11 ’ bUt & “ 6 
m all, and all were carrying load*, \Z & ; there 19 
• e waited for some time 
