STEMONITIS.] STEMONlTACE^. Ill 



minute spines are arranged in the same manner as in a., either giving 

 a close or open reticulation on the surface of the spore or grouped 

 more or less in clusters (not evenly distributed, as in the faintly warted 

 spores of S. splendens). These two varieties represent well-marked 

 centres, but there is no definite boundary between them denoting a 

 true specific difference ; sporangia widely differing in length and with 

 long or short stalks may have spores of either form ; y. occurs with 

 both large and small spores. Rostaflnski's types of S. fusca from Vera 

 Cruz (B. M. 631) and from Ruda Guzowska (Strassb. Herb.) have the 

 spores not smooth, as he describes, but of a form intermediate between 

 a. and |8., 7 to 8 /i diam., with about 28 meshes of reticulation on the 

 hemisphere. S. dictyospora Rost. appears to be an unnecessary name ; 

 it is represented in Kew Herb, by the two types referred to in Rost., 

 Mon., App., p. 27 ; one from Ceylon (K. 1622) bearing the signature 

 of Rostaflnski is S. fusca 0. and has small spores 5 to 6'5 m diam., 

 reticulated in the same manner as in the Strassburg type of S. fusca, 

 but more faintly ; the other type is from Venezuela (K. 1620, B. M. 

 648) on a palm leaf, and in poor condition ; it was marked by Berkeley 

 S. trechispora ; the spores are 10 to 12 ji diam., with a strong complete 

 reticulation in the form of raised bands giving an eveu border to the 

 spore 1 fi broad.' These examples might be taken as representing the 

 extreme Umits in size and reticulation of the spores of is. fusca. The 

 specimen from Venezuela, however, differs so considerably from its 

 nearest allies in the strong and banded reticulation of the spores, that 

 it is a question whether it might not be retained for the present as a 

 distinct species under Berkeley's name of S. trechispora. A type speci- 

 men of 8. maxima Sohwein. received from Dr. Rex of Philadelphia (2697, 

 N. American Fungi, Ellis and Everhart, L:B.M.82), has the spores 7 fi 

 diam., with reticulation precisely of the form above described in 

 Rostaflnski's type of S. fusca in Strassb. Herb. The type of S. 

 nigrescens Rex, kindly furnished by Dr. Bex, has dark spores as in a., 

 but only 7 fi diam. S. Castillensis Macbride, from Nicaragua (B. M. 

 1002) presents no characters by which it can be separated from S. 

 fusca /3. ; the spores are distinctly reticulated, and measure 6 to T fi. 

 The confluent form of the sporangia is in some cases seen throughout 

 the whole development from one plasmodium, the oapillitium con- 

 sisting of a profuse network of arching threads, with broad expansions 

 at the nodes, but sometimes only a papt presents the confluent form, 

 and is associated with more or less perfect sporangia with the normal 

 superficial net. An exceptional form of y. confluens is figured in Plate 

 LXXVII. (L:B.M.82) ; it was found in Epping Forest developing 

 from white plasmodium on dead leaves near rotten wood. The 

 sporangia are combined into a convolute sethalioid mass, the mem- 

 branous sporangium- walls are to a great extent persistent ; no stalks 

 are developed, but in one case the whole ffithalium is suspended by a 

 long slender thread of hypothallus ; the columelte are wanting, and 

 the capillitium is represented by a scanty network of irregular threads 

 with many wide expansions, attached at the extremities to the 

 sporangium - walls. The spores are perfectly formed, 6 fi diam., 

 minutely warted, with the warts here and there connected by faint 

 lines suggesting the appearance of a reticulation. This development 

 is interesting as showing to what extent variation may occur ; if it 

 were not connected with the type with intermediate forms, the position 

 of the specimen might be difficult to determine. The description and 

 figure of AmaurochcBte speciosa Zukal (I.e.) leave little doubt that his 

 species is the form y of S. fusca. 



