KNOWLTON: FIGS FROM WYOMING AND MonTaANA 391 
ance from the interior. In the fossils’ the presence of this pore 
is demonstrable, but it must have been much smaller and not 
provided with anything like a diagram. Curiously enough the 
fossils show little if any evidence of pressure or distortion, which 
is probably to be accounted for by the fact that the beds in which 
they occur are nearly horizontal and exhibit no evidence of having 
been much disturbed since their deposition. They have what was 
probably a thick, woody “shell,” which in all cases appears to have 
been broken across the neck or basal portion just above the point j 
of attachment, thus permitting access to the interior cavity. This 
cavity is now filled with the coarse sandstone of the general matrix, 
mixed with certain extraneous matter, as bits of vegetation, frag- 
ments of shells, etc. In no case was the cavity found to contain 
seeds, which seems rather remarkable considering the fine state of 
preservation of the fruit as a whole. It seems probable that 
when the neck was broken the larger, globose end, being heavier, 
floated downward and the fruits were filled and covered up in 
this upright position in which they are found. 
It should be mentioned that the approximate size of these fruits 
is as follows: length 2 to 3.5 cm., diameter of neck usually a 
little less than 1 cm., and the diameter of the globular portion 
about 2cm. While these sizes are much larger than the average 
living Ficus fruit there are a number of well-known species having 
them even larger, such, for instance, as the Indian F. punctata 
Thunb., F. callicarpa Miq., and others. 
Fruits of Ficus in a fossil state are of unusual occurrence and 
none, so far as I know, has been described that can be removed 
from the matrix practically perfect. Thus Heer in 1874 described 
Ficus protogaea* from the Atane beds of Greenland, and in the 
Dakota group Lesquereux has found two fruits which he figurest 
without specific name. More recently Hollick has described and 
figuredt another species as F. neurocarpa from the Dakota group 
of Ellsworth County, Kansas, and Cockerell§ has established a 
species (F, Bruesi) from the Miocene of Florissant, Colorado. 
* Fl. Foss. Arc. 3?: 108. pl. 30. f. 5-7. 4. 
T Mon. U. S. Geol. Surv. (Fl. Dak. Gr.) 16:-SS. 61. 30. J. 7, 8. r85r. 
¢ Bull. Torrey Club 30: 20S. fo. Fy £903. 
§ Torreya 10: 223. ro10. [Illust.] 
