FOSSIL PLANTS 499 



stone, formed of an outer ring with crenulate borders, and a comparatively 

 large oval or elongated vascular scar in the middle. The double scars evidently 

 represent the point of attachment of single leaves, which, if they had any anal- 

 ogy of form to that of their base, should have been one-third of an inch broad, 

 with round sub- cylindrical borders, and a broad, flat, medial line. The surface 

 of the trunks is regularly and finely wrinkled in the length : the scars trans- 

 versely and still more finely so. The cicatrices, in descending towards the base 

 of the tree, gradually change their form. They first become united into one, 

 forming a deep triangular depression, with a single oval scar at the bottom, 

 and further down in reaching the divisions representing their roots, they be- 

 come round, with a central vascular point, exactly like those of Stigmaria 

 ficoides, though a little smaller. The divisions of the stem, at first inclined 

 downwards, become nearly horizontal at the broken extremities, distant twelve 

 inches from the base of the stem. The largest and best preserved of these 

 trunks is, near its base, four to five inches in circumference, dividing there in 

 nine cylindrical branches, the largest ones seven to nine inches in diameter, 

 merely forked near the broken end, which is two to three inches in diameter. 

 The smallest ones, five inches across, are simple. These divisions, though 

 marked with stigmaroid scars, appear indeed like roots, but it is evident that 

 species of Sigillaria have sometimes grown in sand, and I believe that, under 

 such circumstances, their subterranean divisions have somewhat modified their 

 form, and hence they become similar to roots, as do the stems of Utricularia 

 when they grow in sand. It is <to this kind of organs or roots of Sigillaria, 

 that the fragments described in this report are referable, under the generic 

 name of Sigittarwides* 



From what is said of the relation of Stigmaria with Sigillaria, it is evident 

 that though the forms of Stigmaria are much alike, and generally as yet re- 

 ferred to one species, viz., S. faoides, Brgt., we have indeed as many species of 

 Stigmaria as of Sigillaria, In his Permian, Prof. Goppert still describes Stig- 

 maria ' ficoides with eleven varieties, I cannot see why differences, though diffi- 

 cult to appreciate, should be considered as specific for one genus and as a mere 

 variation for the other. But botanical palaeontology is a peculiar science. It 

 offers to its adepts mere fragments of organs, whose relation to a whole is most- 

 ly unknown, forcing him either to generalize, and to consider in one species a 

 number of organs which evidently pertain to plants of various kinds, or to 

 specify and to divide under divers genera and species, fragments which, if not 

 evidently, at least often apparently, belong to the same vegetable. It is not 

 surprising that opinions concerning these remains are often at variance and 

 often modified, or that the student of these fossil remains becomes discouraged 



*I have never had an opportunity of publishing descriptions with plates of these remarka- 

 ble trees. It may be done hereafter in the report of the Indiana Geological Survey. 



