246 COMPARATIVE MORPHOLOGY OF FUNGI 



age. At a definite time this develops to a linear stroma which is only 

 slightly raised from the dark sclerotium (Fig. 161, 4). 



In the Brazilian B. ambiens, the perithecial outgrowths are more 

 sharply separated from the pseudoparenchymatous stroma. B. ambiens 

 on the roots of Olyra emerges between the leaf sheaths in the forms of 

 a linear stroma 1 mm. broad. The stroma is hard and remains sterile; 

 in solitary irregularly scattered spots, it develops to a short stipitate 

 black heads about 2 mm. thick in which the perithecia are formed (Fig. 



161, 5). 



In a third species, B. redundans, this little head projects above the 

 sclerotic plectenchyma on a stipe 5 mm. long (Fig. 161, 6 and 7). Thus 

 the fertile tissue has attained a definite form while the sterile tissue still 

 has a formless stromatic texture. 



A fourth species, B. diadema, is important for the course of our dicussion, 

 in that it does not parasitize the roots but rather the reproductive organs 

 of the grasses. It surrounds one or two spikelets of a Brazilian Panicum 

 with a comparatively soft stroma whose exterior is darker but not 

 differentiated as a true rind; a thickly interwined hyphal tissue protruding 

 between the glumes, replaces the pistil (Fig. 161, 9). The perithecial 

 heads sprout from this dark stroma, as in B. ambiens (Fig. 161, 8). In a 

 fifth species finally, B. Claviceps, widely dispersed in the tropics on Setaria 

 and Penisetum, the spikelets are surrounded, as in Balansia diadema (Fig. 



162, A) but, in contrast to that species, the stromata are sclerotic as in 

 those of B. ambiens. 



Claviceps forms the end of this series in which the stromata develop 

 to true sclerotia with thick rinds, suited for long resting periods. The 

 separation of the stroma into fertile and sterile parts, lacking in Epi- 

 chloe, barely indicated in Ophiodotis, and leading to a definite form of the 

 fertile part in Balansia, is still sharper in Claviceps; each of the two parts 

 has its characteristic form and develops as its special function requires. 



The simpler species of Claviceps, as the Brazilian C. balansioides, 

 connect directly to Balansia Claviceps. C. balansioides also entangle the 

 flowers and floral organs of grasses (Echinochloa sp.), surrounding the 

 whole tangle with a sclerotic stroma. The enclosed floral organs are 

 retained largely, and the sclerotia are only plumper reproductions of the 

 entangled floral organs. Nevertheless they show considerably greater 

 independence than Balsania diadema and B. Claviceps, for directly after 

 their appearance they no longer proceed to the formation of their peri- 

 thecial heads, but rest for a period of several months. No longer does 

 their tissue merge gradually with the stipe of the perithecial heads, but 

 these soft plectenchymatic fertile parts are laid down under the rind of 

 the sclerotia and become erumpent. 



In the transition from Balansia Claviceps to Claviceps balansioides, 

 the sclerotic plectenchyma fulfils special biological duties, especially of a 



