plate 6, fig. 6) have blunt ends and appear as fragments of a contin¬ 
uous thread. Whether they are fragments that have broken up or are 
separate distinct threads we do not know. Reasoning from analogy 
as concerns the threads of others of the Bovista, and from the fact that 
we find none that are tapering we are inclined to the latter view. 
CLASSIFICATION. 
The Bovistae is composed of three genera, which as far as we 
know, are entirely distinct and do not shade into each other. The 
genus Bovista however, does shade into the genera Bovistella and Ly- 
perdon of the Lycoperdae. 
Capillitium of separate, much branched threads, with slender pointed 
branches,.Bovista. 
Capillitium of separate, entire or fewfbranched, threads, bearing spiny 
points,. . Mycenastrum. 
Capillitium of short, entire, smooth threads with blunt ends, . . Catastoma. 
201—BOVISTA. 
With a little experience the genus is recognizable on sight by 
reason of external appearances, the nature of the peridia and the habits 
of the plants. Of course there are internal features and microscopic 
conditions that confirm the distinction. There may be it is true, plants 
with the “Bovista appearance” outside that have entirely distinct inter¬ 
nal structure, but we do not know of any. Plants are known that have 
the same internal structure, but are different in their habits and nature 
of the peridium, these we call Bovistella (*) (see Myc. Notes, p. 85). 
The peridium of Bovista is firm, parchment-like, elastic, per¬ 
sistent. The plant breaks away from the place of growth, and persist 
tumbling about on the ground for months. There are two layers of 
the peridium in young plants, the outer very thin and called the cor¬ 
tex. As the plant ripens the cortex peels off in patches and entirely 
disappears from the perfectly ripe plant. The cortex of most species 
is smooth, nor does it develop spines as do most species of Lycoperdon. 
Bovista pila has a smooth cortex when very young. As it grows it 
becomes somewhat broken up in scales or granules, but it is never 
“pilose.” 
The capillitium of Bovista consists of sepai ate threads as shown 
in our plates. In the mass, these threads are firm and elastic, and 
thus the spore mass of Bovista can be distinguished by the eye 
from that of other genera of “puff-balls.” This is an adaptation to 
further the method of spore dispersion of Bovista (cfr. Myc. Notes, 
p. 85). The spores of most species of Bovista are pedicellate, a few 
are not. Lycoperdon pedicellatum is the only gastromyces I know 
that has truly pedicellate spores and does not belong to Bovista, or 
the related genus Bovistella. 
(*) Systematists who attach importance entirely to structural characters will call them all 
Bovista. This is simply a question of opinion. I am partial to genera that present features by 
which I know them as soon as I see them. Acknowledging theoreticall } 7 the structural impor¬ 
tance of the distinction, I do not fancy a genus like “Ulocolla” where you have to “sprout the 
spores” before you can tell it. 
114 
