Jennings — Two New Species of Phycopeltis, New Zealand. 757 



Our knowledge of the group is in fact in much the same position as 

 that of older botanists with reference to such genera as Conferva. 

 Classification on characteristics of the vegetative part of the plant only 

 has always proved unsafe. Through the whole range of cryptogamic 

 botany experience has shown that there is so safe ground for the dis- 

 crimination of genera and species outside the study of fully developed 

 plants with their characteristic fructification. 



It may certainly be urged by some systematic botanists that the 

 measurement of cells is of great importance in specific distinctions. It 

 is true that this method has been widely and valuably used in the 

 species of Trentepohlia ; but the conditions of growth in a cell-plate 

 are very different from those in a cell-filam"ent ; and the measurement 

 of cell-dimensions, even if a true average can be struck, is of still 

 less value. 



Moreover it is quite certain that in such species as Ph. tropica, 

 Mob, and in the Ph. nigra of this paper, the same species may appear 

 on the same leaf in several varieties differing only in the size of the 

 cells. 



Following then on the lines of Karsten and Hariot we may thus 

 define the Genus Phycopeltis. 



Trentepohliacese with a thin plate-like thallus of varying shape, 

 consisting of one layer of cells only, growing epiphytically on the 

 surface of leaves, without endophytic rhizoids. True sporangia 

 solitary on a pedicel (commonly uncinate) borne above the disc on 

 a reduced filament of shortened cells. Barren hairs may or may not 

 be present. 



In this generic definition, in addition to the species described by 

 Karsten from Java, and independent of the sterile European forms 

 above referred to, should be included two species collected by the 

 present writer in jSTew Zealand, which may be shortly described as 

 follows. 



A. Phycopeltis expansa, sp. nov. 



Thallus thin, pale yellow, forming at first flabelliform or circular 

 plates, which later on spread into wide expansions, covering a great 

 part of the leaf. Margin always entire, never lobed, and never tend- 

 ing to break into filaments. Cells rectangular, with an average size 

 of -01 X -007 mm. 



Sporangia of two kinds — (a) Disc-sporangia, or enlarged cells 

 intercalated in the cell-rows, not terminal ; (b) solitary sporangia on 



