OBSERVATIONS ON HABITS OF FALL WEBWORM. 43 



specially abundant in that year upon the elm. These and the several 

 facts previously noted in regard to the favorite food plants of the 

 worms seem to indicate that the food plants vary from 3'ear to year 

 and i^erhaps from place to jjlace. It has been suggested that there 

 may be several varieties of the Avorm, but at present I can do no 

 more than call attention to the facts. 



ON WALNUT. 



There are only a few walnut trees at the Point, but the worms 

 played havoc with these; while of all the great abundance of choke- 

 cherry only two instances Avere noted where the infestation was at 

 all similar and so extensive. A clump of five walnut trees (PI. I, 

 fig. 1). each about (> inches in diameter, became literally defoliated 

 and about ir)0 nests were counted upon them. I have observed, 

 however, that the number of nests does not necessarily indicate the 

 number of broods, since a large brood mav desert, its web. divide, and 

 each division form new nests. (See Other Observations.) When 

 food became scarce the worms began to migrate down the trunks of 

 the trees and oA'er the ground, here and there covering the limbs 

 i.nd trunks, and also the ground about the trees, with web. This 

 migration occurred chiefly at night, when the worms literally cov- 

 ered the ground for a radius of several feet about the trees, and all 

 was activity. The migrating worms generally rested, as is custom- 

 ary for them, during the day. In this case they mainly rested in 

 temporary webs often located at the base of the trees. Many of 

 these webs were of extraordinary size and comjwsed of a number of 

 successive sheets of silk arranged parallel with their plane surfaces 

 and about one-fourth inch apart. In this way the capacity of a web 

 was greatly increased, and I estimated that one of the largest con- 

 tained not less than 2 quarts of the worms. 



The worms migrated mainly eastward to a clump of choke-cherry 

 bushes close by and westward to a large hackberry- tree about 40 feet 

 distant. Some webs of migrants were also found in two other hack- 

 berry trees about 50 and 100 feet, respectively, to the southwest and 

 northwest. The migration extended over a period of about ten days, 

 from August 1 to August 10. In four days the number of nests 

 in the chokecherry bushes increased from G to 'lo, and the worms 

 literally stripped the bushes as the}^ advanced. There were only a 

 few native broods upon the choke-cherry bushes. I could alwaj^s 

 distinguish these from migrants by the presence of dried leaves in 

 the nest, there being none of these in the nests of migrants. Then, 

 again, a nest of migrants almost alwaj'S contains worms of various 

 sizes. The absence of dried leaves, together with the presence of 

 worms of several different sizes, I consider to be a verv' good test 

 for distinguishing a migrant nest from a native nest. The advance 



