L YCOPERDALES 



137 



in the various species. The spore-mass is often commingled 

 with thread-like bodies either simple or intricately branched, 

 known as the capillitium ; this is commonly attached to the 

 columella or walls, but in some genera it is free. The outer por- 

 tions often separate into two or three layers and collectively are 

 known as the peridia. The outer peridium in the ordinary puff- 

 ball separates in small flakes from the inner and falls away at or 

 before the maturity of the spores. In the earth- stars it splits into 

 a series of teeth and becomes flattened or more commonly re- 

 flexed. In some forms both layers of the peridium break up into 

 a series of fragments and the spores become scattered either by 

 the wind or more rapidly by the accidental treading of some ani- 

 mal. In others the inner peridium opens by an apical mouth 

 and when compressed belches forth a puff of smoke-like spores. 

 The genus Catastoma (PI. g. f. J-/) presents a curious anomaly 

 in spore dispersion which was long misunderstood until its true 

 method was brought to light by Mr. Morgan. The outer peridium 

 raptures in a circumscissile manner about the equator of the puff- 

 ball its upper portion remaining attached to the apical half of the 

 ball, while the ball itself becomes separated from the lower por- 

 tion which remains in the ground as a cup. The ball then be- 

 coming overturned, opens with a mouth on its original under 

 surface or basal portion, and like the species of some other genera 

 becomes free to be rolled about by the wind, and thus scatters its 

 spores over a wider area. Myriostouia ( PL 7. /. 7 ), a rare but 

 interesting species,* opens by a series of perforations so that the 

 top resembles that of a pepper-box. Some species have a colu- 

 mella formed of a portion of the stem extending up into the spore- 

 chamber ; in a few genera this columella extends through to the 

 apex of the peridium ; in rare cases the spore cavities are sepa- 

 rated by radiating lamellae and thus a series of apparent con- 

 necting links unite the puff-balls to the agarics. Thus far a num- 

 ber of peculiar or unique forms have appeared on the Pacific 

 coast and in the semi-arid regions of New Mexico and Arizona, 

 but little is yet known except of the mere existence of these 

 genera. The knowledge of the development, affinities and dis- 



* This has been reported from only three stations widely separated, in 

 Colorado, Florida and Ontario. Cf. Morgan, Am. Nat. 26: 341, 342, 

 1892 ; and Cook, Bot. Gaz. 23 : 43, 44, 1897. 



