116 Rhodora [May 



tions of the leaf -cells and the underleaves (see Figs. 14, 23-25). In 

 the case of the female inflorescence, however, a few remarks may be 

 in place. Although the perichaetial bracts are almost invariably 

 reduced to a single pair, a second pair is occasionally developed. 

 When this is the case the pair next to the perianth is essentially the 

 same as when a single pair is present. Midler implies that there is a 

 much greater discrepancy in size between the lobe and lobule than in 

 P. pUiiyphyUa, but this idea is not borne out by the writer's observa- 

 tions. As a matter of fact the bracts (Figs. 33, 34, 36, 37) are much 

 the same as in P. phdyphylla and are subject to the same variations 

 in the character of the apex and the degree of the marginal dentation; 

 sometimes, indeed, the apex is tipped with an apiculum (Figs. 33, 34). 

 In the case of the bracteole teeth are occasionally present in the upper 

 part as well as in the basal region (Figs. 35, 38). 



The mouth of the perianth is more contracted in P. plafyphylloidca 

 than in P. platyphyUa and is densely ciliate. The cilia are usually 

 one cell wide and vary from one to fi\e or more cells in length (as 

 Schiffner states) ; sometimes compound cilia are present with second- 

 ary cilia along their sides, but simple unbranched cilia are the rule. 

 The cilia are sometimes straight as in Fig. 40 but it is much more 

 usual for them to be variously contorted as in Fig. 39. Projecting 

 cells in the neighborhood of the mouth are sometimes present as in the 

 North American specimens of P. platyphyUa (see Fig. 39). 



In studying a large series of capsules of P. platyphyUoidca the elaters 

 have shown for the most part a single spiral throughout their length. 

 Elaters with two spirals in the middle portion have, nevertheless, 

 been repeatedly olwerved, and in some capsules a rather high percent- 

 age of the elaters have been of this type. In no case, however, has an 

 elater with two spirals throughout its entire length been noted, so that 

 in the elaters at least the differences between P. platyphyUa and P. 

 platyphyUoidca are definite and constant. In distinguishing the two 

 species the most trustworthy differences are, in fact, those drawn from 

 the elaters. Boulay ^ comments on the difficulty of using sporophytic 

 characters in the genus at all on account of the habitual sterility of 

 material, but this difficulty does not detract in any way from the 

 importance of such characters. 



In the absence of mature capsules the mouth of the perianth, even 

 when undeveloped perianths are the only ones available, should be 



1 Muse, de la France 2: 18. 1904. 



