198 PROC. ENT. SOC. WASH., VOL. 22, NO. 8, NOV., 1920 



to keep it fresh, and well stoppered to prevent the larvae from 

 falling in and drowning. A wide-mouthed No. 2 lantern globe 

 covered with gauze at the top was placed over each twig to pre- 

 vent escape. The water and leaves were changed once a week. 

 The larvae were looked over three times a week, the excrement 

 removed, and necessary measurements taken. The length of the 

 caterpillar was measured by simply using a millimeter rule, ex- 

 cept during the first instar. The length during this time, and 

 the width of the head and body at all times were measured with 

 a binocular microscope and micrometer. This instrument was 

 accurate to 0.03 mm. To obtain these measurements upon live 

 moving caterpillars was very trying upon the observer's eyesight 

 and patience. After each molt the cast heads were placed in a 

 small vial at the base of the twig, thus furnishing a check upon 

 the measurements and number of molts. These cast heads can 

 be measured, furnishing the head measurements in an easier way 

 than taking them while on live larvae. 



The mortality of the caterpillars was very great, particularly 

 during their early life. Consequently, to obtain satisfactory data, 

 a large number of young caterpillars had to be started in rearing 

 and many measurements taken, which later proved of little use, 

 due to the death of those individuals. Only a relatively small 

 number lived to be pupae. Of these, eleven normal individuals 

 furnished averages for most of the foregoing measurements. 



Feeding Habits. 



The small larva, upon hatching, makes its first meal upon the 

 egg shell. Next it starts feeding upon the epidermis of the leaf 

 near to the place of hatching, and is joined by its comrades, twenty 

 to forty strong. They usually feed gregariously in this manner 

 throughout the first instar, leaving a considerable area with the 

 surface removed and a rough brown network of veins beneath. 



During the second instar the larvae separate and then begin 

 feeding upon the full thickness of the leaves, and in heavy infes- 

 tations eating all except a few of the larger veins before changing 

 to pupae. 



Migration. 



A number of experiments were performed in order to discover 

 how far newly hatched larvae would travel in search of food. 

 Larvae were released at certain points very soon after having 

 eaten their egg shells. In fifteen minutes some were starting to 

 roam about, and in an hour several were from ten to twenty- 

 three inches away. Four and one-half hours after being released, 

 several were from fifty to eighty-three inches distant. Others 

 did not start to roam until the second and third days and most 



