756 



SGIENGE. 



LN. S. Vol. III. No. 73. 



Tisual netted impressions characteristic 

 and familiar in the fossil remains. The 

 leaf scar bears three well-marked points. 

 The Lepidodendra always branch dichoto- 

 mously. 



Concerning the mode of development of 

 the primary xylem or central part of the 

 stele there has been lack of evidence and 

 consequent radical difference of opinion. 

 To the solution of this problem Prof. "Wil- 

 liamson devoted the six summer months 

 of 1894 examining the slides in his ex- 

 traordinary collection numbering several 

 thousand specimens, and counting or cal- 

 culating with great precision the number of 

 cells in the primary and secondary xylem 

 systems. 



The study of the dichotomies of the 

 branches has thrown great light upon this 

 important question, for "It is to the as- 

 cending series of these dichotomies that 

 the Lepidodendra owe their characteristic 

 structure and modes of development." 



The first change in the normally cylin- 

 drical ordinary branch is the splitting of 

 the central vascular cylinder of primary 

 xylem and its contained cellular medulla. 

 The cylinder splits vertically for a short 

 distance into two crescentic, diverging 

 halves, while the external form of the 

 branch becomes oval, the diiference between 

 the longer and shorter axes being greater 

 as we ascend to the dichotomy. Before 

 reaching this the two horns of each cres- 

 cent of primary xylem approach each 

 other. At this stage the cells of the two 

 m^edullse are in direct contact with those of 

 the inner cortex. Several of Prof. Wil- 

 liamson's sections show that the crescentic 

 condition is permanent for a short space at 

 least, approximating what DuBary calls 

 the ' foliar gap ' in ferns. Higher in the 

 dichotomy, however, the horns of each 

 crescent rapidly converge to form two new 

 cylinders, ' differing in no respects, save 

 size and number of internal parts, from 



that of the parent stem.' The same phe- 

 nomena occur with each successive dicho- 

 tomy, each pair of resulting branches, 

 though diminished in diameter, having 

 exactly the same type of organization as the 

 one from which they sprang. Thus, how- 

 ever numerous the dichotomies they are all 

 produced alike and no structural changes 

 are introduced ' from the base of the 

 parent branch up to the smallest twig 

 of the full-grown tree,' save certain second- 

 ary ones produced by growth processes 

 which begin to manifest themselves at 

 the base of that trunk. It will at once 

 be seen that the number of cells and 

 vessels of the cortex, primary xylem and 

 medulla is one-half as great in each of the 

 two branches as in the parent below the 

 bifurcation. A similar ratio obtains in the 

 number of leaves. Such dichotomies occur 

 only when the twigs are terminal with a 

 growing point. 



Besides these equal dichotomies there are 

 two sorts of unequal dichotomies different 

 in structure and pui'pose. In the first only 

 a small segment is cut out of the primary 

 xylem cylinder and passes outward, carry- 

 ing with it a small portion of the medulla 

 to form a branch. This unsymmetrical 

 segment becomes a solid strand with or 

 without any trace of a medulla, though 

 usualljr on reaching the axis of the cone, 

 which it usually supports, a central medulla 

 is shown. In the second form of unequal 

 dichotomy the medulla is unaffected, a 

 limited number of tracheids being detached 

 from the periphery of the primary xylem 

 cylinder. These strands may also go to 

 reproductive organs. 



In answer to the hitherto open problem 

 as to how far the ordinary growth of a 

 branch has exerted any influence upon or 

 borne any relation to the varying dimen- 

 sions of the primary xylem cylinder of the 

 stele and upon the number of its com- 

 ponent tracheids, the author's examination 



