PENNSYLVANIA TREE FERN COMPRESSIONS 15 



the way that the living ones grow, all of these patterns could occur 

 in one species. 



Apparently, natural species of tree ferns cannot be distinguished 

 by leaf scar patterns. However, form-species can be useful and should 

 be clearly defined. The limits for defining a form-species should be 

 large enough to include a wide variation of characteristics but small 

 enough to show any trends of change that might be present. The following 

 characteristics are suggested for use and are used in this paper: 

 1) the height to width ratio of the leaf scar; 2) the size (height) of 

 the leaf scar; and, to a certain degree, 3) the shape of the leaf scar. 



The different configurations of the vascular bundle found in 

 leaf scars of Artisophyton can be interpreted as a developmental 

 sequence in one species (text fig. 4). The specimen shown on plate 4 

 demonstrates that different configurations can and did occur on one stem. 

 The suggestion that different configurations occur in the same species 

 was first made by Corsin (1948). 



The species differentiation in Artisophyton must thus be based 

 on artificial characteristics (size and height to width ratio) the same 

 as those used in the other stem genera. The differentiation is somewhat 

 simpler in Artisophyton because there are only two types of leaf scars, 

 those with a height to width ratio greater than 1 (height > width) and 

 those with a ratio of 1 and less than 1 (height < width) . 



Orientation of Leaf Scars 



Most of the leaf scars of Caulopteris, Megaphyton, and Artiso- 

 phyton are found as fragments of stems and thus do not allow any conclu- 

 sions about the proper orientation (adaxial side versus abaxial side) . 

 Therefore, figures in the literature are oriented either way without 

 regard to the natural orientation on the living stem. 



A horseshoe-shaped vascular bundle occurs in Caulopteris , Mega- 

 phyton, and Hagiophyton. If we assume that these three genera were simi- 

 lar in organization to Psaronius and the living Marattiaceae, the proper 

 orientation is defined. The horseshoe is adaxially (at the top) open; and 

 in the closed stage (see text fig. 10, E and F) , the inner trace is in the 

 upper portion of the outer trace. Unless the contrary is proven, all leaf 

 scars of these three genera should be shown in that orientation. 



The orientation of the leaf scars of Artisophyton is more prob- 

 lematical because no comparable living or petrified plant is known. The 

 original orientation of the stem cannot be determined from single leaf 

 scars that occur in ironstone nodules. There is, however, a specimen 

 (ISGS PB-1298) in which both sides of an Artisophyton stem are preserved 

 (pi. 4). A reconstruction is shown in text figure 13. The row of scars 

 to the left represents the o'utside of the stem facing the viewer. On 

 the right we look into the stem, the internal tissue being gone and re- 

 placed by a thin shale layer. Thus the row of leaf scars on the right 

 is the compression of the outer layer of the stem as it appears from 

 the inside. This information alone does not yet tell us the natural 

 position of the stem. Some data derived from the vascular bundle must 

 be added. The preserved surfaces not only show the vascular bundle 



