FLORA OP THE GALAPAGOS ISLANDS. 97 



Chatham Isl. : March, 1884, Marcacci, ace. to Piccone, 1. c. Further 

 distrib. western So. Am. 



Herpophyllon, gen. prov. 



Frons membranacea, prostrata, subcircularis vel vague expansa, 

 centro radicibus adfixa. Structura interne cellulis oblongis vel rudia- 

 tim elongatis, parietibus distincte colloideis cellulis superficiem versus 

 rotundatis in stratum unicum corticale transformatis constituta. Spor- 

 angia cruciatim divisa, in soris verrucaeformibus supra frondem sparsa. 

 Cjstocarpia ignota. 



H. coalescens, spec. prov. 



Frondes circa 40 mm. diametro, aggregatae, a marginibus coalescentes, 

 thallum indefinite expansum formautes. Sporangia in soris verrucae- 

 formibus ad superficiem superiorem frondis inter paraphyses allata, an- 

 guste ovales, 56 /i X 15-20 /n. Albemarle Isl.: Tagus Cove, March, 

 1899, Snodgrass & Heller. 



The ^provisional name given above is adopted to designate a curious 

 alofa the cystocarpic fruit of which is unknown and which in other respects 

 does not agree with any genus known to me. When removed from the 

 fluid in which it was preserved it appeared like an irregular, rather 

 cartilaginous pellicle adherent to the substratum by groups of coarse 

 rhizoids. On a closer examination the surface was seen to be ridged, 

 or veined, and apparently the larger specimens are composed of smaller 

 individual fronds which have united, each tuft of rhizoids being in the 

 centre of a frond. The smaller specimens were like the larger, but none 

 seen was so small as to be composed of a single frond. The microscopic 

 structure of the ridges show that they are the lines of union of two different 

 fronds, or, if the whole is to be regarded as a single plant, of its different 

 lobes. The internal structure consists of a compact tissue of large cells 

 about 5G/itin transverse section but more or less elongated in radial 

 sections, some being 140 /x long. They cannot, however, be called 

 filaments. Near the upper and lower surfaces and near the margin 

 the cells become more nearly spherical. The cell-walls are markedly col- 

 loidal and near the margin seem to be imbeded in a solid colloidal matrix. 

 The cortical cells are small and arranged in a single layer, especially on 

 the lower surface, those on the upper surface more frequently dividing 

 anticlinically, especially where the sporidia are forming. The latter are 

 in spots on the upi)er surface, which are sometimes not much raised 

 but at other times form well-marked warts. The sporangia are borue 

 between parapliyses about GU /i long and 3 ft broad, formed by the out- 



VOL. XXXVIII. — 7 



