STEMONITIS.] STEMONITACE.E. 113 



L. cribrosus (Syst. Myc., iii., p. 87) implies that he probably had the 

 confluent form of a Stemonitis before him. S. confluens Cke. & Ellis, 

 from New Jersey, Ellis (K. 665 ; and L:B.M.83, part of the same 

 gathering, furnished by Dr. Rex), appears also to be a confluent form 

 of S. sphndens ; the spores in both the N. Carolina and New Jersey 

 specimens have the typical sculpture, but are darker than usual, and 

 measure 9 to 10 /j. diam. A specimen from Meudon in the collection of 

 the Paris Museum closely resembles that from New Jersey in the 

 character of the capillitium ; the spores have also the same dark tint, 

 and measure 10 to 11 /n ; but the sporangia are more normal, having in 

 some cases a simple columella and a nearly complete superficial net 

 with a wide mesh. Only three or four European gatherings of this 

 species are represented in the Strassburg, Brit. Mus., and Kew Collec- 

 tions ; it is plentiful in India, America, Australia, and the Pacific 

 Islands, from which regions there are numerous specimens in the 

 collections, which were classed under S. fusca, until Rostafinski 

 detected the specific characters and gave the name of S. sphndens. 

 The capillitium in this species exhibits wide differences, but the spores 

 are remarkably constant in colour, size, and in the minute, evenly 

 distributed warts, which are sometimes scarcely apparent, even when 

 magnified 1,200 diam. ; their distribution resembles that on the spores 

 of Physarum nutans. The superficial net of the capillitium appears 

 to be continuous with the evanescent sporangium-wall, which is not 

 merely attached by short spines projecting from the net as in S. fusca; 

 this character is illustrated by a remarkable form described by Dr. 

 Rex (Proc. Acad. Nat. Sc. Phil., 1890, p. 36) under the name 

 S. Bauerlimi Mass., /. fenestrata. He records the appearance of suc- 

 cessive growths of the Stemonitis at considerable intervals of time, on a 

 limited area of a decaying log, apparently from one original source. 

 Through the courtesy of Dr. Rex the gatherings are represented in the 

 mountings in the Brit. Mus. In mounting (a) the sporangium-wall 

 is persistent except in approximate circular perforations 10 to 20 p, wide, 

 or in other words the superficial net is expanded to form a perforated 

 wall to the sporangium. Mounting (b) is from a later gathering, with 

 much of the character of (a), but approaching nearer to the normal 

 form. Mounting (c) is from a crop appearing a month later than (5), 

 in which there is a still more marked return to the usual habit, with 

 the meshes of the net 30 to 60 p, wide. The width of the mesh varies 

 in Rostafinski's types from Cuba and Texas (referred to Rost., App., 

 p. 27) ; in that from Cuba (B. M. 630) the average width of the mesh 

 is 70 p, in that from Texas (K. 1631) it is 20 ju. S. Morgani Peck, 

 N. Am. Fungi, Ellis & Everh. 2088, and S. Bauerlimi Mass., from 

 New Guinea (K. 726), are essentially the same form as the Cuba type, 

 the mesh of the superficial net averaging about 60 p in width, 

 S. Webberi Rex (/. /3) has a wider mesh than the Cuba type, and is 

 described (I.e. 1891, p. 391) as distinguished from S. sphndens by the 

 spores being ferruginous-coloured in mass, and by the pale surface 

 capillitium ; the mounted specimens do not show this difference of 

 colour. The form gathered at Lyme Regis in 1891 (Journ. Bot. 1891, 

 p. 263), var. y, has even more lax and broken capillitium than var. /3, 

 and the spores in mass are rich purple-brown ; the growth has appeared 

 on the same fir stumps in abundance in 1892 and 1893, with much the 

 same characters as in the first gathering. It has also been obtained 

 from the New Forest, Hants, from the Black Forest near Freiburg, 

 and from Ohio. The type specimen of S. acuminaia Mass. (K. 698) 

 is a. genuina^ the spores measuring 7 to 8 p. diam. In looking through 



8 



