156 MISSOURI BOTANICAL GARDEN. 
this is the cause of the bursting of the bands was determ- 
ined repeatedly by noting that in no case’ would the bands 
burst until a considerable formation of wood took place in 
the twig to which the bags were attached. The bursting of 
the bands took place at almost any point in their circum- 
ference except at the point where the bag was attached to 
the band; in other words, the weakest point in the bands 
usually was at some point away from the bag itself. This 
was a fact of considerable interest, the force of which will 
be referred to again further on. 
FORMATION OF THE SWELLINGS. 
By the end of July practically all of the bands of the 
previous year have dropped from the trees; here and there 
however, one finds abag which has not fallen. Upon ex- 
amination it will be found that the band of this particular 
bag still passes intact around its twig; and should this bag 
remain upon the twig until September, one will find that a 
swelling, such as the ones referred to above, has begun 
to form. It would appear from this that any particular 
band which remains throughout the summer has been strong 
enough to resist the forces exerted by the growing twig as 
it expanded in the formation of new wood, and that the 
band then acted as a ligature. All of the conditions for 
the formation of the swelling were thereby brought about. 
As the bag worm occurs on other trees besides the 
arbor vitae, an examination was made of trees in the 
Garden and in the streets of St. Louis for the purpose of 
finding out whether the girdling action noted for the 
arbor vitae trees also took place on other trees. After 
a few days of searching it was found that practically every 
tree species upon which the bag worm occurred was girdled 
now and then very much as were the arbor vitae trees. On 
some species the girdling action appeared to bevery common, 
particularly on the sycamore and soft maple, but nowhere 
did it occur as frequently as with the arbor vitae. Thus 
