WHITE AMTS AliTD FUNGI. .191 



towards the exterior are directed radially, and all these 

 terminate ^t first in spherical or oval swellings. Some of 

 thom produce single spherical cells, 25 to 40 ^ in diameter, 

 which form to some extent an outer covering to the sphere ; 

 these cells are at fi.rst terminal, butbecome lateral bysubsequent 

 growth of the hypha. Others produce a chain of three to six 

 oval cells of varying size, and then revert to normal hj^hae. 

 In rare cases, branching occurs in these chains of spore-like 

 cells in the same manner as in the spheres of the termite comb. 



When the agaric develops the spheres turn yellow and 

 collapse. The large spherical cells are then indistinguishable 

 when the sphere is viewed as a solid object under a low 

 magnification, and it appears as simply a clump of ordinary 

 hyphse. The agaric is not formed in the interior of a sphere, 

 but develops on the top of a cluster of them. 



At first sight these spheres appear identical \rith the white 

 spherical bodies which grow on the combs of certain termites, 

 and which Berkeley described under the name of j^gerifa 

 DutJiiei. But closer examination shows that although it is 

 possible to trace some resemblance between the constituent 

 parts of each, they differ widely in the degree of differentiation 

 to which those parts have attained, and completely in the 

 arrangement of them. The Entoloma sphere is a tangle with 

 some approach to a definite arrangement at the exterior, while 

 Mgerita Dutliiei, on the other hand, resembles a true conidial 

 fructification in being composed of distinct branches radiating 

 from a common stalk, the outer of which form branching chains 

 of spherical cells, while the mner form similarly branching 

 chains of regular narrow-oval cells. The definite structure 

 and arrangement of Mgerita Duthiei are entirely lacking in the 

 Entoloma sphere. 



I found this species m April, 1905, growing in profusion on 

 the side ol a mound of earth at the base of a clump of palms. 

 Part of the mound was occupied by a termite nest, and the 

 r3mainder most probably consisted of the debris of previous 

 nests, but at the time the soil was quite loose (not cemented 

 together) and dark coloured, and was covered with grass and 

 other vegetation. Heavy rains had washed away the surface of 

 the moimd and exposed the masses of spheres, which completely 



