446 University of California PuMic-aiions in Botany [Vor.. 6 



the California coast and extends as far north as the Strait of Juan 

 de Fuca. It is not at all unlikely that it may be found on the same 

 host in other localities. 



R. clavata departs slightly from the previous species, R. laniinariae, 

 in having the filaments occasionally branched, and the cross cell 

 divisions often decidedly irregularly oblique. 



Radaisia subimmersa Setchell et Gardner sp. nov. 

 Plate 37, figs. 12-13 



Thallis parvulis, inconspicuis, in linea exteriore in superificie hostis 

 irregularibus, cuticulam aut cavernas externas pai'vasque incolentibus ; 

 thallo prostrato cellulis irregularibus irregulariter positis 3-5/i. diani.. 

 primo verticalibus, tardiore ad fines liberos distantes divergentibus, 

 35-45/x longis, composito ; partitione cellularum in superficie libera 

 thalli in directione una intus in directionibus tribus, pondera ovalia 

 formante ; cellulis filamentorum erectorum 3-5/x longis, 1.5-2.5ju, latis, 

 vetustioribus frequenter pyriformibus ; contentu homogeneo, caeruleo- 

 viridi ; gonidangiis terminalibus, sphaericis, 4-6|ti diam., gonidiis 6-8, 

 partitione simultanea. 



Thalli small, inconspicuous, irregular in outline on the surface of 

 the host, growing on the cuticle, or in small surface cavities ; prostrate 

 or basal layer composed of angular cells 3-5/* diam., arranged irregu- 

 larly, giving rise to erect filaments, parallel at first, later spreading 

 somewhat at the free distal ends, 35-45/a long ; cell division on the free 

 surface portion in one plane, within the host in three planes, building 

 up oval masses ; cells in erect filaments 3-5/x long, 1.5-2.5jLt wide, older 

 cells frequently becoming pear-shaped ; contents homogeneous, blue- 

 green; gonidangia terminal, spherical, 4-6/a diam., producing 6-8 

 gonidia by simultaneous division. 



Growing on RJwdymcnia sp. Carmel Bay, Monterey County, 

 California. 



The host plant, no. 3350, was collected by me. It is a Rhodymenia 

 and possibly R. palmetta, at least belongs to the palmetta group. It 

 was cast ashore and slightly faded thus making the Radmsia groups 

 appear distinct which led to their discovery. The host is not uncom- 

 mon along the California coast and the epiphyte is to be expected in 

 other localities. This species of Radaisia, like others, seems to be 

 epiphytic at least until the basal layer of cells is produced, as shown 

 in plate 37, figure 13. 



Plants start from a single cell on the cuticle of the host, at least 

 numbers of them have been so observed. By repeated divisions, some- 

 times perpendicular to the longer diameter of the cell, but usually 

 quite oblique, a single layer of cells of varying shapes and sizes is 



