316 Mr. W. Fawcett on an Entomogenous Fungus. 



thinner and pass over directly into the central fibrils. The 

 processes of the ganglion-cells of the second type are broad 

 dark fibres, upon which the sheath extends its cells for a long 

 distance. They can therefore be easily traced in the nervous 

 system among the fine fibrils, especially as their breadth does 

 not diminish. After running some distance they lose their 

 sheath and soon afterwards disappear in the mass of fine 

 fibrils. It seems to me most probable that, like the colossal 

 nerve-fibres, which they greatly resemble, tliey pass into the 

 fibrillar substance by brush-like division, as I could never 

 observe any binary division. 



If we examine the ganglion-cells of the second type, and 

 especially the colossal ones, we find that the whole cell is 

 traversed in all directions by fibrils of different strengths, 

 which pass over into the cell-process and give it a fine longi- 

 tudinal striation. But these fibrils do not quit the cells only 

 in this way ; one is astonished to see how they issue every- 

 where at the periphery of the naked cell-body, singly or united 

 into bundles, and penetrate into the subcuticular envelope. 

 This observation may be made uniformly in preparations 

 hardened in alcohol or in corrosive sublimate or osmic acid. 

 Whether the ganglion-cells are united to one anotlier by means 

 of these fibrils I have been unable to decide, as they were not 

 to be traced beyond the envelope of subcuticular fibres. 



XXXII. — An Entomogenous Fungus. 

 By William Fawcett, B.Sc, F.'L.S. 



Dr. GiJNTHER has received from Mr. C. A. Lloyd, George 

 Town, Demerara, a remarkable fungus growing on an ant 

 {Camponofus atriceps). Mr. Lloyd found it on the banks of 

 the river Puruni, in British Guiana, and though he has col- 

 lected numerous specimens of different species of ants, he has 

 not hitherto met with a similar growth. It is not so usual to 

 find them on a perfect insect as on larvae. Of the 47 species 

 noted by Saccardo *, 23 (or about 50 per cent.) are found on 

 larv£e, and only 16 (or about 33 per cent.) on perfect insects. 

 Of these 16 species Saccardo only mentions 3 as having been 

 found on ants — Cordyceps unilateralis on Atta ceplialoia in 

 Brazil, C. australis on Pachycondgla striata in Brazil, and 

 0. myrmecophila on Myrmica rufa (also rarely on an ichneu- 



* ' Sylloge Fungorum,' vol. ii. 



