I06 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



diameter, slightly constricted at the septa towards apex, scarcely 

 so below; cells 1-3 (mostly 1.5-2) times as broad as long, the 

 apical broadly dome-shaped or subhemispheric; heterocysts sub- 

 spheric or ovoid, basilar, rarely double, 5.0-6.6 ix in diameter, or 

 sometimes 8.3 ix long; spores unknown. [Plate 12, figures 

 12-17.] 



On Dictyota dichotoma, dredged "in 133^-14 fathoms," Lewis 

 Radcliffe, August 11, 1914; accompanied by Streblonema soli- 

 tarium, Elachistea stellulata, Phaeostroma pusillum, Acrochaetium 

 affine, etc. 



It was at first suspected that this plant might prove to be 

 identifiable with Microchaete purpurea J. Schmidt,^ described as 

 an epiphyte on species of Fiiciis dredged near the Danish island 

 of Laes0 in the Kattegat, but a comparison with type material 

 of this species kindly furnished by Dr. F. B0rgesen of Copenhagen 

 showed important differences. The trichomata of Microchaete 

 nana are shorter than those of M. purpurea and are 5,0-8.3 jx in 

 diameter vs. 3-5 /x in M. purpurea; the filaments are loosely 

 gregarious and are mostly decumbent or ascending, while those 

 of M. purpurea form densely congested, suberect, wick-like clusters, 

 which in turn form on the surface of its host subdendroid or fucoid 

 figures, slightly suggestive of certain forms of frost-crystals on a 

 window-pane; the vaginae of the filaments of M. nana are thinner, 

 more delicate and less perceptible, though our material of M. nana 

 is formalin-preserved, while that of AI. purpurea is dried, which 

 may account for a certain amount of the difference in this respect; 

 the color of the trichomata of M. nana is, perhaps, not well pre- 

 served, but it would seem to be a light olivaceous rather than the 

 red-purple of M. purpurea. Possibly Microchaete vitiensis Aske- 

 nasy, described from islands in the Pacific Ocean, is a nearer rela- 

 tive. rVom description alone, M. vitiensis would appear to have 

 erect and longer filaments ("fills millimctrum vix attingentibus") 

 with thicker sheaths, which finally become ocrcate. 



Derbesia turbinata sp. nov. 



Olive-green and nitcnt when dry, more or less repent, apparently 

 forming straggling mats 8-9 cm. broad (or high?), the basal j^arts 

 sometimes here and there resoKed into cysts; filaments 15-95 )li 

 (mostly 38-53 ix) in diameter, sparingly branched, the branching 



' Bot. Ti(]sskr. 22: 379, 412. iSyy. 



