his studies on root disease in Hawaii. It was again mentioned by 

 Cobb (15) in a succeeding Hawaiian bulletin. In neither instance- 

 Avas it named, although a possible connection wdth Marasmius sacchari 

 was suggested. Bancroft (7) in British Guiana in reporting on a 

 root disease which he attributed to ^larasmius, mentions stellate 

 crystals as one of the characteristics of the mj'celium. What is 

 apparently the same fungus was reported from Porto Ricn l)y Faw- 

 cett (26), from St. Croix by Longfield Smith (69, 70) and from 

 Jamaica by Ash by (3). 



The presence of this fungus on other plants than cane suggests 

 an error in the current method of combating root disease, the turn- 

 ing of the land to pasture. It will be much more effective to plant 

 a legume (eowpeas or velvet beans), and so starve out the fungus. 



Himantia stellifera Johnston sp. nov. 



Mycelium cob-webby, or somewhat dentritic, white, ascending the 

 lower leaf-sheaths and penetrating the roots. Hyphae with clamp 

 conections, and bearing on short side l)ranches stellate crystals of 

 calcium oxalate. No fruiting bodies known. 



Differs from Himantia sacchari Speg. and Himantia guttulifera 

 Speg. in the presence of the stellate crystals as well as in other essen- 

 tial points. 



Porto Rico. — On cane, Rio Picdras, Jan. 12, 1914, 1183, (type) ; 

 Juncos, Aug. 23; 1915, 3050; Cambalache, March, 1914, 5136. On 

 Cymbopogon cih-atus, Rio Piedras, Oct. 26, 1914, 2282, Oct. 1915, 

 3212. Collected or observed on cane and many other hosts commonly 

 in all parts of Porto Rico. (PI. XIX, fig. 2; p1. XXXI, figs. 1-4.) 



XHE GRANULAR LEAF-SHEATII MJNGUS {Odonim sdrchaHcola) . 



This is a very common fungus occurring at tlic base of cane stalks 

 l>inding, (as is usual with fungi of this nature) the lower leaf-sheaths 

 firmly together and to the stalk itself as well, so as to make their 

 removal difficult. Whether or not tlie fnii<ins ix-nctratcs hdow 

 ground to the extent of attacking the roots is a point which it has 

 not yet been possible to investigate. The only visible (himagc is the 

 rotting of the leaf-sheaths whieh are permeated by the wliitc mycelial 

 elements. Fruiting areas occur as uniform thin white patches with 

 a granular somewhat poAvdery surface encircling the stalk from the 

 ground level to a lieight of eight inches or a foot. (I'l. XTX, fig. 1.^ 



This fungus is very commonly found on all varieties of cane and 

 occurs apparently independent of whether the stool is liealthy or 

 otherwise. It is very commoii on canes attacked l)y the mottling 



188 



