REPORT OF THE DIRECTOR, 1922 59 
scattered sclerenchyma strands may be due entirely to some macera- 
tion before preservation; but the ring itself appears to be a definite 
zone, part of the missing central cylinder. Transverse sections of 
larger trunks were made, but nothing was found. Success in this 
line, I believe, can only be attained when we find a petiole or rachis 
of a frond preserved in such a condition that thin sections can be 
made for study. 
Roots. Though we have a number of specimens of roots, none 
of the stumps taken out have any roots attached. For this reason 
it has been suggested that the stumps perhaps may not have grown 
in the place in which they were found. There is one trunk now, 
still in place in the quarry at Gilboa at this time of writing, which 
shows the roots spreading outward in the shale layer (lowest hori- 
zon) in which the stumps rest; roots were likewise observed in the 
shale layer of the highest horizon from which stumps were taken. 
These observations taken into consideration with the upright position 
of all the stumps found, the successive tree horizons with the bases 
of the stumps in each case resting in and upon a shale layer, repre- 
senting the clay bed in which the trees are believed to have grown, 
would all seem to indicate that the forest was destroyed and the 
stumps buried im situ. 
Eospermatopteris apparently had no large heavy roots. 
None have been found and there certainly is nothing about the 
stumps themselves that would suggest that any might be expected. 
All the roots found, in their flattened condition, measure between 
one quarter of an inch or less and half an inch across; but they are 
not preserved in a condition which permits of their being studied 
strueturally. Numerous long, strap like rootlets, as shown in text 
figure 2 and plate 8, radiate from the roots, but do not appear to have 
any regular arrangement. These rootlets, as preserved, measure on 
the average 1% of an inch or less across and there is little, if any, 
variation in width. Some of the rootlets measure up to 6 inches in 
length as far as preserved, and there is no indications that this is 
anywhere near the tip. With a root system such as this, I think the 
bases of these trees must have been buried to some height in the 
sediments to give adequate foothold. 
Foliage. The fronds of Eospermatopteris are tripin- 
nate; they were at least 6 feet long, and, judging from the frag- 
ments of main rachises collected, in the larger specimens they must 
have been considerably longer. One specimen (figure 3) shows an 
impression of a main rachis 25 inches long which is apparently at 
