1955] 
Brooks — Fossil Galls 
3 
perpaniculata from Graves County, Kentucky, with embay- 
ments cut into its margins, 6 to 9 mm. across, is figured 
and discussed. The notches are crudely semicircular and 
are variable in outline and in size. Berry admits the 
dubious nature of these injuries and suggests as an alter- 
native that the cuts may have been made by caddis fly 
larvae for the construction of cases. 
These examples are all suggestive of the work of leaf- 
eating insects. The last specimen cited may have been cut 
by a primitive leaf-cutting bee, the work of which appears 
cruder than that of the modern Megachilidae but can hardly 
be regarded as conclusive evidence of the presence of 
megachilid bees in the Eocene forest of Tennessee and 
adjoining states. 
A bee-cut N ectandra pseudocoriacea leaf, more authentic 
than those described by Berry, was collected by R. Lee 
Collins at Puryear, Tennessee. The leaf has four semi- 
circular pieces, 5 to 6 mm. in diameter, cut from one of 
its edges (pi. 1, fig. 1 and la). The darkened borders 
of the wounds appear to be lignitized remains of scar 
tissue. Recent female megachilid bees inflict similar in- 
juries in cutting pieces of leaves for the construction of 
nests. Semicircular pieces are cut for the construction of 
the sides and circular pieces for partitions. In contrast, 
damages to leaves produced by phytophagous insects are 
irregular in shape. The scar tissue proves that the leaf 
remained on the tree for some time after it was injured 
and could not have been cut by a caddis fly larva for pieces 
of leaves to construct a case (pi. 1, fig. 6) as Berry sug- 
gested (1931, p. 303). The uniformity of the shape and 
size of the injuries and the presence of the scar tissue is 
credible evidence that the fossil leaf was damaged by a 
leaf-cutting bee. 
No fossil Megachilidae have been reported from rocks 
older than the Oligocene deposits of Florissant, Colorado 
(Cockerell, 1908). The deposit has also yielded a leaf 
with semicircular notches cut into its margin which Cock- 
erell (1910, p. 429) states is evidence that the habit of 
cutting out pieces of leaves for building nests was as highly 
developed by the megachilid bees during the Oligocene as 
