42 PENH ALLOW AND DAWSON 



film already referred to. Similar thiu coaly films also fill the narrow clefts which 

 traverse the stem at somewhat frequent intervals in a longitudinal direction, and which 

 obviously originated in alteration under pressure. 



An examination of the outer coaly layer from No. 5 shows that it has exactly the same 

 structure as the interior silicified portions, whence we may infer that, on the one hand 

 there could have been no difi'erentiation of a proper cortical structure, and that the tissue 

 of the stem was of the same kind throughout ; or else that with removal of the bark, were 

 it originally present, the exposed surface layers of tissue were converted into coal. 



Finally, we are prepared to consider one more structural peculiarity which has a 

 most important bearing upon the true character of bur plant, and its systematic position. 

 "We have already called attention to the fact that the large tubular cells take a horizontal 

 or obliqiie direction, chiefly in the vicinity of the radial openings, into which they even- 

 tually run. The frequency with which this was found to occur, led us to the belief that 

 there must be some special connection between the cells themselves and the open areas 

 which they penetrated. For some time no adequate explanation could be reached, but 

 after a careful search through all the sections then on hand, a single large cell entering 

 the open area in a horizontal direction, was found to be distinctly branched. It had pre- 

 viously been noted that on entering these openings, the large cells usually became 

 thinner-walled and more attenuated. The cell in question, however, was found to terminate 

 in two distinct branches, the ultimate ramifications of which could not be determined on 

 account of their having been cut away in making the section. In more recently prepared 

 sections, we have found several such branching cells. In one case, five branches were 

 found to arise from a cell at a section which was hardly longer than the diameter of the 

 cell, and where there was a strong contraction in the latter. Other instances have also been 

 noted. Not unfrequently the branching is indicated simply by a small round hole in the 

 cell wall, where the branch projected towards the observer. In Plate II, fig. 5, on the 

 extreme lower left-hand margin, there may be seen a cell with five distinct branches ; 

 while to the right of the centre there is another cell with at least seven well-defined 

 branches, some of which project towards the observer, and, therefore, appear as round 

 openings only. 



This branching, so far as observed, occurs only in the radial openings, though we 

 see no reason why it may not occur elsewhere. From the relations which we have thus 

 established, we have no hesitation in stating that the system of smaller, intercellular 

 myceloid filaments are nothing more nor less than the ultimate ramifications of the larger 

 cells, and that the radial openings are the special areas within which the branching is 

 accomplished. 



In order to gain a clearer idea as to whether or not the stru^cture described could be 

 considered normal, we examined a number of sections of fossil woods, in which the oper- 

 ation of decay was too evident to admit of question. "We also submitted to examinatioLi 

 an old tamarack [Larix Americana) water-log, which, having been in use for eighty years, 

 was assumed to represent, to some extent at least, conditions of decay which would be 

 liable to be represented in fossil woods. In neither case could we obtain any evidence 

 to show that the fossil now under consideration had suiFered special decay ; indeed, all 

 the evidence pointed rather strongly in the opposite direction. It was at one time sup- 

 posed by us that the intercellular filaments were fungoid mycelia incident to decay, but 



