that there must have been a grass understory. Members 
of the Ericaceae, especially Andromeda, must also have 
occurred in the understory immediately under the pines, 
because flowers, fruits and leaves were all preserved in 
the resin. Schubert (1958) emphasized that the dry con- 
dition of the forest was particularly indicated by the nar- 
row annual rings in many of the small branches of pine. 
Czeczott (1960), however, agreed with Anders that the 
Florida hammock-type forest is more xerophytic than is 
indicated by both the overall nature of the amber flora 
and many members of the insect fauna. They believe that 
the mixed oak-pine forest was dense and moist. Other 
deciduous trees occurred on the outskirts of the oak-pine 
forest and in open glades. The more tropical plants lived 
on the southern slopes of an area that must have been 
partly mountainous. On the basis of additional wood and 
bark studies, Schubert (1961) emphasized that there was 
not just one forest-type, but that there probably were 
pine-palm, pine-oak, and pine-hardwood types forming a 
mosaic pattern on mountain slopes. He holds, nonethe- 
less, that his studies (Schubert, 1958; 1961) support an 
hypothesis of a warm-dry woods of a savannah-type. He 
compares the amber forest to the situations now occurring 
in mountainous areas in Cuba (Lotschert, 1957), as well 
as possibly in southern Mexico, Guatemala, and British 
Honduras (Martinez, 1948; Loock, 1950; Schwerdt- 
feger, 1953). 
It was suggested by Heer (1860, 1869) and followed 
by Tornquist (1910) that the amber forests covered an 
extensive area. ‘Tornquist thought that the southern 
boundary of the forest extended across what is now cen- 
tral Sweden eastward through Finland into western 
Russia, whereas the adjacent northern sea occupied not 
only what is now northern Germany but also the region 
drained by the Vistula and Dnieper Rivers as far as the 
[ 251 J 
