104 Rhodora [May 



can, in which the elaters were bispiral througliout, showed that some 

 of the gametophytic characters rehed upon by Miiller were subject 

 to variation, although his characterization of the species was essen- 

 tially correct. In considering these variations the lobes, the lobules, 

 the underleaves, the bracts and bracteoles, and the mouth of the peri- 

 anth will be taken up in order. The leaves and underleaves discussed 

 are in all cases those of the stem itself. 



Although the lobes are normally ovate as described (see Figs. 1, 2, 

 4, 5) they are sometimes as broad as long and may be practically 

 orbicular in outline (Figs. 3, 6). The cordate condition of the base, 

 moreover, is not always as clear as might be desired. At the junction 

 with the short keel, to be sure, the base is distinctly dilated into a 

 cordate expansion, which may be either plane, or slightly crispate or 

 revolute, but this expansion is by no means as constant on the other 

 side of the lobe. In the case of the lobes shown in Figs. 4 and 5 the 

 basal dilations are distinct enough, and similar dilations are shown 

 in Figs. 1 and 2; but in Figs. 3 and 6 the lobe gradually narrows toward 

 the base and is actually short-decurrent. It should be noted in this 

 connection that the line of attachment of the lobe is strongly arched, 

 so that the extent of the decurrent portion is not easy to define. The 

 margin of the lobe is not quite so uniformly entire as ]\Iiiller's descrip- 

 tion implies, and the portion of the margin which extends from the 

 dorsal base to the apex is frequently more or less dentate. The teeth, 

 which in some cases at least mark the position of slime papillae, are 

 exceedingly irregular in size, in number and in distribution, and they 

 are usually little more than blunt and \aguely defined projections 

 (Figs. 1, 2, 5). Sometimes, however, they are sharper, and it is not 

 unusual to find one or two sharp and minute teeth close to the base 

 (Figs. 1, 2). Aside from these basal teeth, a larger and very variable 

 tooth about midway between the base and the apex is fairly frequent. 

 The presence of teeth on the lobe is briefly alluded to by Warnstorf.^ 



In the case of the lol)ule the line of attachment is even more sharply 

 arched than in the case of the lol)e and makes the degree of decurrence 

 even more difficult to define clearly. Perhaps the most accurate 

 method would l)e to consider the lobule decurrent only when the inner 

 edge extends farther backward than the outer edge, and this is appar- 

 ently what is implied in Midler's description. His account of the 



' Kryptogamenflorader Mark Brandenburg 1: 270. 1902. 



