262 PETCH : THE KlNCl 



balls of mycelium were not identical with the termite spheres, 

 but as there was a termite nest in close proximity, the question 

 of a possible connection was left open. 



Unfortunately what should have been the wet season of 

 1905 was abnormally dry. and I did not find the agaric again 

 until May 1. 1906. The rainfall for April was only 4*5 in., 

 but 2 *85 in. fell on the last five days. This time there was no 

 doubt that the agaric was independent of termite nests. It 

 occurred in various places on the flower beds, and in great 

 abundance on a bank by the roadside. In the latter locality 

 the mycelium was again exposed and some had been washed 

 Iry the rain into the gutter, where it developed agarics among 

 the dead leaves. The largest clump of mycelium formed a 

 thin cake IS cms. long and 6 cms. broad, but having been 

 exposed was spoilt by the attacks of insects, especially a pod- 

 urid. Ackarutes armatus, Nicolet. The mycelium appears to 

 lie generally at a depth of one or two centimetres and therefore 

 when growing on a slope is soon laid bare by the rain. It seems 

 probable that it is formed in this position at the beginning of 

 the rains and is at once exhausted by the growth of the agarics. 

 There are no strands of mycelium connecting adjacent clumps. 



The spheres are - 4 to "7 mm. in diameter and are massed 

 together like fish roe : they lie in a compact mass, without any 

 vegetable matter among them, and are bound together by 

 fine hypha- which run from each sphere to all others in contact 

 With it: some of these hypha 1 are united into strands like 

 those which occur in the mycelium on the termite comb. 



The interior of a sphere is a tangle of interlacing hypha 1 

 swollen into irregularly oblong or oval cells, which may be 

 produced in succession, or singly in the course of a hypha. 

 There is no definite arrangemenl of the individual hypha? in 

 the interior, hut radially-directed hypha? near the eiivumfer- 

 enoe of t be sphere show more regularity, and all these terminate 

 at final in Bpherieal or Oval swellings (fig. .*W ). Some of these 

 form single Bpherioal oelk 25 to Lfl /. diam. which simulate a 

 covering on the exterior • fchey are borne singly on the hypha and 



