CALOSTOMATACEAE 187 



easily broken off, and after dehiscence they are usually represented only by dark col- 

 lapsed lines over the region of attachment, giving this part a roughish appearance. 



Microscopic structure of the peridia is as follows: the outer fibrous layer is com- 

 posed of small and densely packed threads, which in the outer region run parallel to the 

 surface and curving through a right angle continue through the horny layer perpen- 

 dicular to the surface. In this last layer the threads are straight, almost perfectly parallel 

 to each other and very densely packed; continuing into the spongy layer they at once 

 become looser and much more entangled. The thin inner layer of the spore sac is not 

 vcrv different from the spongy layer, and from its surface the threads composing it pass 

 into the gleba without obvious change to form the capillitium. 



Maire makes the peculiar statement that the basidia are terminated by a "sort of 

 single compound sterigma at the top of which the spores are borne on very short pedi- 

 cels." He gets the number of spores right. Tulasne gives the basidial shape right, 

 but the number of spores as four. As they are of considerable interest, we quote below 

 in full Maire's observations on this species (Comp. Rend. 131 (2): 1248. 1900): 



"Gcaster hygromctricus shows basidia with 2-8 spores. The fusion nucleus under- 

 goes two successive divisions; sometimes the divisions stop there, and there are formed 4 

 spores, each of which receives a nucleus. At other times one or two of the 4 nuclei 

 divided again, from which result 5 spores and 5 nuclei or 6 spores and 6 nuclei; often 

 the 4 nuclei divide and there result 8 spores and 8 nuclei. Sometimes the nucleus of the 

 basidium undergoes only one division, whence two nuclei and two spores. The most 

 usual case seems to be the 6-spored basidium. On the other hand, the formation of the 

 spores is, up to a certain point, independent of that of the nuclei: it happens, for 

 example, that a basidium having only 4 nuclei forms 5 spores, etc. These processes 

 are very variable, quite as in the basidia of the Cantharellaceae which we are studying 

 at this time. However that may be, in Geaster, the formation of the spores always 

 takes place in the following manner: at the tip of the basidium is sent out a single 

 prolongation, a sort of compound sterigma, at the summit of which the spores are 

 formed on very short pedicels. The nuclei, of definite number, are drawn all together 

 into the compound sterigma, to be separated afterwards, each in its spore. In Geaster, 

 the centrosomes and the kinoplasm are scarcely visible." 



For other cytological information by Maire, covering this and other species, see 

 under Scleroderma auranlium. 



Up to the present the attachment of this plant to the soil has been completely 

 misunderstood. This is hard to understand as the basal rhizomorphs are so conspicu- 

 ous. DeBary's figure (Fungi, Myc. and Bact, fig. 146) shows no basal attachment 

 and he says the button is attached to a felted mycelium which often spreads in the soil 

 for the distance of an inch all around. This description completely reverses the facts 

 and the error has never been corrected. The delicate flocculence of the surface of the 

 buttons is so little connected with the soil that they come up almost clean. This is 

 entirely different from the intimate attachment to the soil of the mycelial surface layer 

 of hypogeal Geasters. It will be noted that this is an additional character pointing 

 to a relationship with Colostoma and Tylosloma. 



We have examined the type of Geastrum fibrillosum in the Schweinitz Herbarium, 

 and find that it is, as Lloyd says (Geastrae, p. 10), a weathered specimen of this species. 

 It is the flaky roughness of the outer surface that Schweinitz referred to in his name, 



