232 PETCH : THE FUNGI 



and had branched into the tree-like structure photographed. 



Thus from the bend upwards represents fourteen hours' growth. 

 The former whip-like apex is seen to the right of the bend. The 

 bell jar was removed for about twenty seconds: during that 

 time the branches collapsed slightly and the looser stromata 

 commenced to droop, as shown on the light of the photograph, 

 They doubled over completely immediately the photograph 

 had been taken. On the following day nearly all had assumed 

 the tree-like form and the stalks had turned black. The deli- 

 rate radiating hyphse are not well seen in the photograph (1*1. 

 XIII.C): the stroma on the extreme left especially shows their 

 collapse. The main stalk hyphffl were about 12^ diameter with 

 frequent septa, some only 30 ju apart: some of the ultimate 

 branches were only 5 // in diameter. This cultivation was 

 continued until 17th November, but no conidiophores were 

 observed. The mass of spreading hyphsB on the comb is chiefly 

 .i Minor. 



The corresponding pot cultivation will be dealt with later. 

 It gave results which have not been continued by any Othei 

 experiment, but which agree exactly with those obtained by 

 observation of the combs in the nest. 



A diminution of moisture results in the formation of simi- 

 larly branched stromata in which the branches are rigid, taper 



to a point, and bear scattered conidiophores along their whole 



Leilgl h. These conidiophores branch repeated I \ . and are identi- 

 cal with those oi figS. 29 and .".<». 



It is possible when the comb is produoing simple stromata 



only, to cause the formation of branched ones with rigid ooni- 

 diferotts branches bj the addition of water. This was done in 



one instance with a semi-desert ed romb. i.e. it contained work- 

 oid soldiers only. It u;is placed under a bell jar on 23rd 



Februarj without removing all the termites These were 



removed as the\ appeared beneath the comb, and during that 



tune all developing fungi were eaten. The last two termites 



ken out on 1st March. \ development of simple stro- 

 mata iimu took place and by Ith .March these were about 1 



